Merge branch 'sensor-interface#200' into pointCloud
This commit is contained in:
Néstor Subirón
GitHub
commit
66af651711
|
|
@ -1,9 +1,11 @@
|
|||
Dist
|
||||
Doxygen
|
||||
PythonClient/dist
|
||||
Util/Build
|
||||
|
||||
*.VC.db
|
||||
*.VC.opendb
|
||||
*.egg-info
|
||||
*.kdev4
|
||||
*.log
|
||||
*.pb.cc
|
||||
|
|
@ -22,5 +24,7 @@ Util/Build
|
|||
.tags*
|
||||
.vs
|
||||
__pycache__
|
||||
_benchmarks_results
|
||||
_images*
|
||||
_out
|
||||
core
|
||||
|
|
|
|||
|
|
@ -24,4 +24,4 @@ matrix:
|
|||
packages:
|
||||
- cppcheck
|
||||
script:
|
||||
- cppcheck Unreal/CarlaUE4/Source Unreal/CarlaUE4/Plugins/Carla/Source Util/ -iUtil/Build -iUtil/CarlaServer/source/carla/server/carla_server.pb.cc --quiet --error-exitcode=1 --enable=warning
|
||||
- cppcheck . -iBuild -i.pb.cc --error-exitcode=1 --enable=warning --quiet
|
||||
|
|
|
|||
|
|
@ -5,18 +5,19 @@
|
|||
"path": ".",
|
||||
"folder_exclude_patterns":
|
||||
[
|
||||
"__pycache__",
|
||||
".clang",
|
||||
".codelite",
|
||||
".kdev4",
|
||||
".vs",
|
||||
"Build",
|
||||
"Binaries",
|
||||
"Unreal/CarlaUE4/Content*",
|
||||
"Build",
|
||||
"DerivedDataCache",
|
||||
"Dist",
|
||||
"Doxygen",
|
||||
"Intermediate",
|
||||
"Saved"
|
||||
"Saved",
|
||||
"Unreal/CarlaUE4/Content*",
|
||||
"__pycache__"
|
||||
],
|
||||
"file_exclude_patterns":
|
||||
[
|
||||
|
|
@ -47,6 +48,18 @@
|
|||
},
|
||||
"build_systems":
|
||||
[
|
||||
{
|
||||
"name": "CARLA - Pylint",
|
||||
"working_dir": "${project_path}",
|
||||
"file_regex": "^\\[([^:]*):([0-9]+):?([0-9]+)?\\]:? (.*)$",
|
||||
"shell_cmd": "pylint --disable=R,C --rcfile=PythonClient/.pylintrc PythonClient/carla PythonClient/*.py --msg-template='[{path}:{line:3d}:{column}]: {msg_id} {msg}'"
|
||||
},
|
||||
{
|
||||
"name": "CARLA - CppCheck",
|
||||
"working_dir": "${project_path}",
|
||||
"file_regex": "^\\[([^:]*):([0-9]+):?([0-9]+)?\\]:? (.*)$",
|
||||
"shell_cmd": "cppcheck . -iBuild -i.pb.cc --error-exitcode=0 --enable=warning --quiet"
|
||||
},
|
||||
{
|
||||
"name": "CARLA - Rebuild script",
|
||||
"working_dir": "${project_path}",
|
||||
|
|
|
|||
20
CHANGELOG.md
20
CHANGELOG.md
|
|
@ -1,3 +1,21 @@
|
|||
## CARLA 0.7.1
|
||||
|
||||
* New Python API module: Benchmark
|
||||
- Defines a set of tasks and conditions to test a certain agent
|
||||
- Contains a starting benchmark, CoRL2017
|
||||
- Contains Agent Class: Interface for benchmarking AIs
|
||||
* New Python API module: Basic Planner (Temporary Hack)
|
||||
- Provide routes for the agent
|
||||
- Contains AStar module to find the shortest route
|
||||
* Other Python API improvements
|
||||
- Converter class to convert between Unreal world and map units
|
||||
- Metrics module to summarize benchmark results
|
||||
* Send vehicle's roll, pitch, and yaw to client (orientation is now deprecated)
|
||||
* New RoutePlanner class for assigning fixed routes to autopilot (IntersectionEntrance has been removed)
|
||||
* Create a random engine for each vehicle, which greatly improves repeatability
|
||||
* Add option to skip content download in Setup.sh
|
||||
* Few small fixes to the city assets
|
||||
|
||||
## CARLA 0.7.0
|
||||
|
||||
* New Python client API
|
||||
|
|
@ -9,7 +27,7 @@
|
|||
- Better documentation
|
||||
- Protocol: renamed "ai_control" to "autopilot_control"
|
||||
- Merged testing client
|
||||
- Added the maps for both cities, the client can now access the car position within the lane.
|
||||
- Added the maps for both cities, the client can now access the car position within the lane
|
||||
* Make CARLA start without client by default
|
||||
* Added wind effect to some trees and plants
|
||||
* Improvements to the existing weather presets
|
||||
|
|
|
|||
|
|
@ -0,0 +1,73 @@
|
|||
# Contributor Covenant Code of Conduct
|
||||
|
||||
## Our Pledge
|
||||
|
||||
In the interest of fostering an open and welcoming environment, we as
|
||||
contributors and maintainers pledge to making participation in our project and
|
||||
our community a harassment-free experience for everyone, regardless of age, body
|
||||
size, disability, ethnicity, gender identity and expression, level of experience,
|
||||
education, socio-economic status, nationality, personal appearance, race,
|
||||
religion, or sexual identity and orientation.
|
||||
|
||||
## Our Standards
|
||||
|
||||
Examples of behavior that contributes to creating a positive environment
|
||||
include:
|
||||
|
||||
* Using welcoming and inclusive language
|
||||
* Being respectful of differing viewpoints and experiences
|
||||
* Gracefully accepting constructive criticism
|
||||
* Focusing on what is best for the community
|
||||
* Showing empathy towards other community members
|
||||
|
||||
Examples of unacceptable behavior by participants include:
|
||||
|
||||
* The use of sexualized language or imagery and unwelcome sexual attention or
|
||||
advances
|
||||
* Trolling, insulting/derogatory comments, and personal or political attacks
|
||||
* Public or private harassment
|
||||
* Publishing others' private information, such as a physical or electronic
|
||||
address, without explicit permission
|
||||
* Other conduct which could reasonably be considered inappropriate in a
|
||||
professional setting
|
||||
|
||||
## Our Responsibilities
|
||||
|
||||
Project maintainers are responsible for clarifying the standards of acceptable
|
||||
behavior and are expected to take appropriate and fair corrective action in
|
||||
response to any instances of unacceptable behavior.
|
||||
|
||||
Project maintainers have the right and responsibility to remove, edit, or
|
||||
reject comments, commits, code, wiki edits, issues, and other contributions
|
||||
that are not aligned to this Code of Conduct, or to ban temporarily or
|
||||
permanently any contributor for other behaviors that they deem inappropriate,
|
||||
threatening, offensive, or harmful.
|
||||
|
||||
## Scope
|
||||
|
||||
This Code of Conduct applies both within project spaces and in public spaces
|
||||
when an individual is representing the project or its community. Examples of
|
||||
representing a project or community include using an official project e-mail
|
||||
address, posting via an official social media account, or acting as an appointed
|
||||
representative at an online or offline event. Representation of a project may be
|
||||
further defined and clarified by project maintainers.
|
||||
|
||||
## Enforcement
|
||||
|
||||
Instances of abusive, harassing, or otherwise unacceptable behavior may be
|
||||
reported by contacting the project team at [INSERT EMAIL ADDRESS]. All
|
||||
complaints will be reviewed and investigated and will result in a response that
|
||||
is deemed necessary and appropriate to the circumstances. The project team is
|
||||
obligated to maintain confidentiality with regard to the reporter of an incident.
|
||||
Further details of specific enforcement policies may be posted separately.
|
||||
|
||||
Project maintainers who do not follow or enforce the Code of Conduct in good
|
||||
faith may face temporary or permanent repercussions as determined by other
|
||||
members of the project's leadership.
|
||||
|
||||
## Attribution
|
||||
|
||||
This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 1.4,
|
||||
available at https://www.contributor-covenant.org/version/1/4/code-of-conduct.html
|
||||
|
||||
[homepage]: https://www.contributor-covenant.org
|
||||
|
|
@ -1,23 +1,25 @@
|
|||
Contributing to CARLA
|
||||
=====================
|
||||
|
||||
> _This document is a work in progress and might be incomplete._
|
||||
|
||||
We are more than happy to accept contributions!
|
||||
|
||||
How can I contribute?
|
||||
|
||||
* Reporting bugs
|
||||
* Feature requests
|
||||
* Improving documentation
|
||||
* Code contributions
|
||||
|
||||
Reporting bugs
|
||||
--------------
|
||||
|
||||
Use our [issue section](issueslink) on GitHub. Please check before that the
|
||||
issue is not already reported.
|
||||
Use our [issue section][issueslink] on GitHub. Please check before that the
|
||||
issue is not already reported, and make sure you have read our
|
||||
[Documentation][docslink] and [FAQ][faqlink].
|
||||
|
||||
[issueslink]: https://github.com/carla-simulator/carla/issues
|
||||
[docslink]: http://carla.readthedocs.io
|
||||
[faqlink]: http://carla.readthedocs.io/en/latest/faq/
|
||||
|
||||
Feature requests
|
||||
----------------
|
||||
|
|
@ -28,6 +30,25 @@ your request as a new issue.
|
|||
|
||||
[frlink]: https://github.com/carla-simulator/carla/issues?q=is%3Aissue+is%3Aopen+label%3A%22feature+request%22
|
||||
|
||||
Improving documentation
|
||||
-----------------------
|
||||
|
||||
If you feel something is missing in the documentation, please don't hesitate to
|
||||
open an issue to let us know. Even better, if you think you can improve it
|
||||
yourself, it would be a great contribution to the community!
|
||||
|
||||
We build our documentation with [MkDocs](http://www.mkdocs.org/) based on the
|
||||
Markdown files inside the "Docs" folder. You can either directly modify them on
|
||||
GitHub or locally in your machine.
|
||||
|
||||
Once you are done with your changes, please submit a pull-request.
|
||||
|
||||
**TIP:** You can build and serve it locally by running `mkdocs` in the project's
|
||||
main folder
|
||||
|
||||
$ sudo pip install mkdocs
|
||||
$ mkdocs serve
|
||||
|
||||
Code contributions
|
||||
------------------
|
||||
|
||||
|
|
@ -42,6 +63,17 @@ please contact one of us (or send an email to carla.simulator@gmail.com).
|
|||
|
||||
[projectslink]: https://github.com/carla-simulator/carla/projects/1
|
||||
|
||||
#### Where can I learn more about Unreal Engine?
|
||||
|
||||
A basic introduction to C++ programming with UE4 can be found at Unreal's
|
||||
[C++ Programming Tutorials][ue4tutorials]. Then, if you want to dive into UE4
|
||||
C++ development there are few paying options online. The
|
||||
[Unreal C++ Course at Udemy][ue4course] it's pretty complete and there are
|
||||
usually offers that make it very affordable.
|
||||
|
||||
[ue4tutorials]: https://docs.unrealengine.com/latest/INT/Programming/Tutorials/
|
||||
[ue4course]: https://www.udemy.com/unrealcourse/
|
||||
|
||||
#### What should I know before I get started?
|
||||
|
||||
Check out the ["CARLA Design"](carla_design.md) document to get an idea on the
|
||||
|
|
@ -50,36 +82,56 @@ the new feature. We are aware the developers documentation is still scarce,
|
|||
please ask us in case of doubt, and of course don't hesitate to improve the
|
||||
current documentation if you feel confident enough.
|
||||
|
||||
#### Are there any examples in CARLA to see how Unreal programming works?
|
||||
|
||||
You can find an example of how C++ classes work in UE4 in
|
||||
[`ASceneCaptureToDiskCamera`][capturelink] (and its parent class
|
||||
`ASceneCaptureCamera`). This class creates an actor that can be dropped into the
|
||||
scene. In the editor, type _"Scene Capture To Disk"_ in the Modes tab, and drag
|
||||
and drop the actor into the scene. Now searching for its detail tab you can find
|
||||
all the `UPROPERTY` members reflected. This shows the basic mechanism to use C++
|
||||
classes in Unreal Editor.
|
||||
|
||||
For a more advanced example on how to extend C++ classes with blueprints, you
|
||||
can take a look at the _"VehicleSpawner"_ blueprint. It derives from the C++
|
||||
class `AVehicleSpawnerBase`. The C++ class decides where and when it spawns a
|
||||
vehicle, then calls the function `SpawnVehicle()`, which is implemented in the
|
||||
blueprint. The blueprint then decides model and color of the vehicle being
|
||||
spawned. Note that the blueprint can call back C++ functions, for instance for
|
||||
getting the random engine. This way there is a back-and-forth communication
|
||||
between C++ code and blueprints.
|
||||
|
||||
[capturelink]: https://github.com/carla-simulator/carla/blob/master/Unreal/CarlaUE4/Plugins/Carla/Source/Carla/SceneCaptureToDiskCamera.h
|
||||
|
||||
#### Coding standard
|
||||
|
||||
Please follow the current coding style when submitting new code.
|
||||
|
||||
###### General
|
||||
|
||||
* Use spaces, not tabs.
|
||||
* Avoid adding trailing whitespace as it creates noise in the diffs.
|
||||
* Comments should not exceed 80 columns, code may exceed this limit a bit in rare occasions if it results in clearer code.
|
||||
|
||||
###### Python
|
||||
|
||||
* All code must be compatible with Python 2.7, 3.5, and 3.6.
|
||||
* [Pylint](https://www.pylint.org/) should not give any error or warning (few exceptions apply with external classes like `numpy`, see our `.pylintrc`).
|
||||
* Python code follows [PEP8 style guide](https://www.python.org/dev/peps/pep-0008/) (use `autopep8` whenever possible).
|
||||
|
||||
###### C++
|
||||
|
||||
* Compilation should not give any error or warning (`clang++ -Wall -Wextra -std=C++14`).
|
||||
* Unreal C++ code (CarlaUE4 and Carla plugin) follow the [Unreal Engine's Coding Standard](https://docs.unrealengine.com/latest/INT/Programming/Development/CodingStandard/) with the exception of using spaces instead of tabs.
|
||||
* CarlaServer uses [Google's style guide](https://google.github.io/styleguide/cppguide.html).
|
||||
Please follow the current [coding standard](coding_standard.md) when submitting
|
||||
new code.
|
||||
|
||||
#### Pull-requests
|
||||
|
||||
Once you think your contribution is ready to be added to CARLA, please submit a
|
||||
pull-request to the `dev` branch.
|
||||
pull-request.
|
||||
|
||||
Try to be as descriptive as possible when filling the pull-request description.
|
||||
Adding images and gifs may help people to understand your changes or new
|
||||
features.
|
||||
|
||||
Please note that there are some checks that the new code is required to pass
|
||||
before we can do the merge. The checks are automatically run by the continuous
|
||||
integration system, you will see a green tick mark if all the checks succeeded.
|
||||
If you see a red mark, please correct your code accordingly.
|
||||
|
||||
###### Checklist
|
||||
|
||||
<!--
|
||||
If you modify this list please keep it up-to-date with pull_request_template.md
|
||||
-->
|
||||
|
||||
- [ ] Your branch is up-to-date with the `master` branch and tested with latest changes
|
||||
- [ ] Extended the README / documentation, if necessary
|
||||
- [ ] Code compiles correctly
|
||||
- [ ] All tests passing
|
||||
- [ ] `make check`
|
||||
- [ ] `pylint --disable=R,C --rcfile=PythonClient/.pylintrc PythonClient/carla PythonClient/*.py`
|
||||
- [ ] `cppcheck . -iBuild -i.pb.cc --enable=warning`
|
||||
|
|
|
|||
|
|
@ -1,4 +1,6 @@
|
|||
; Example of settings file for CARLA.
|
||||
;
|
||||
; Use it with `./CarlaUE4.sh -carla-settings=Path/To/This/File`.
|
||||
|
||||
[CARLA/Server]
|
||||
; If set to false, a mock controller will be used instead of waiting for a real
|
||||
|
|
|
|||
|
|
@ -0,0 +1,66 @@
|
|||
CARLA Benchmark
|
||||
===============
|
||||
|
||||
Running the Benchmark
|
||||
---------------------
|
||||
|
||||
The "carla" api provides a basic benchmarking system, that allows making several
|
||||
tests on a certain agent. We already provide the same benchmark used in the CoRL
|
||||
2017 paper. By running this benchmark you can compare the results of your agent
|
||||
to the results obtained by the agents show in the paper.
|
||||
|
||||
|
||||
Besides the requirements of the CARLA client, the benchmark package also needs
|
||||
the future package
|
||||
|
||||
$ sudo pip install future
|
||||
|
||||
By running the benchmark a default agent that just go straight will be tested.
|
||||
To run the benchmark you need a server running. For a default localhost server
|
||||
on port 2000, to run the benchmark you just need to run
|
||||
|
||||
$ ./run_benchmark.py
|
||||
|
||||
or
|
||||
|
||||
$ python run_benchmark.py
|
||||
|
||||
Run the help command to see options available
|
||||
|
||||
$ ./run_benchmark.py --help
|
||||
|
||||
Benchmarking your Agent
|
||||
---------------------
|
||||
The benchmark works by calling three lines of code
|
||||
|
||||
corl = CoRL2017(city_name=args.city_name, name_to_save=args.log_name)
|
||||
agent = Manual(args.city_name)
|
||||
results = corl.benchmark_agent(agent, client)
|
||||
|
||||
This is excerpt is executed in the [run_benchmark.py](https://github.com/carla-simulator/carla/blob/master/PythonClient/run_benchmark.py) example.
|
||||
|
||||
First a *benchmark* object is defined, for this case, a CoRL2017 benchmark. This is object is used to benchmark a certain Agent. <br>
|
||||
On the second line of our sample code, there is an object of a Manual class instanced. This class inherited an Agent base class
|
||||
that is used by the *benchmark* object.
|
||||
To be benchmarked, an Agent subclass must redefine the *run_step* function as it is done in the following excerpt:
|
||||
|
||||
def run_step(self, measurements, sensor_data, target):
|
||||
"""
|
||||
Function to run a control step in the CARLA vehicle.
|
||||
:param measurements: object of the Measurements type
|
||||
:param sensor_data: images list object
|
||||
:param target: target position of Transform type
|
||||
:return: an object of the control type.
|
||||
"""
|
||||
control = VehicleControl()
|
||||
control.throttle = 0.9
|
||||
return control
|
||||
The function receives measurements from the world, sensor data and a target position. With this, the function must return a control to the car, *i.e.* steering value, throttle value, brake value, etc.
|
||||
|
||||
The [measurements](measurements.md), [target](measurements.md), [sensor_data](cameras_and_sensors.md) and [control](measurements.md) types are described on the documentation.
|
||||
|
||||
|
||||
|
||||
Creating your Benchmark
|
||||
---------------------
|
||||
Tutorial to be added
|
||||
|
|
@ -9,7 +9,7 @@ This document describes the details of the different cameras/sensors currently
|
|||
available as well as the resulting images produced by them.
|
||||
|
||||
Although we plan to extend the sensor suite of CARLA in the near future, at the
|
||||
moment there are three different sensors available. These three sensors are
|
||||
moment there are only three different sensors available. These three sensors are
|
||||
implemented as different post-processing effects applied to scene capture
|
||||
cameras.
|
||||
|
||||
|
|
@ -38,6 +38,8 @@ more human readable palette of colors. It can be found at
|
|||
Scene final
|
||||
-----------
|
||||
|
||||
<br>
|
||||
|
||||
The "scene final" camera provides a view of the scene after applying some
|
||||
post-processing effects to create a more realistic feel. These are actually
|
||||
stored on the Level, in an actor called [PostProcessVolume][postprolink] and not
|
||||
|
|
@ -55,6 +57,8 @@ in the Camera. We use the following post process effects:
|
|||
Depth map
|
||||
---------
|
||||
|
||||
|
||||
|
||||
The "depth map" camera provides an image with 24 bit floating precision point
|
||||
codified in the 3 channels of the RGB color space. The order from less to more
|
||||
significant bytes is R -> G -> B.
|
||||
|
|
@ -81,12 +85,14 @@ Our max render distance (far) is 1km.
|
|||
Semantic segmentation
|
||||
---------------------
|
||||
|
||||
|
||||
|
||||
The "semantic segmentation" camera classifies every object in the view by
|
||||
displaying it in a different color according to the object class. E.g.,
|
||||
pedestrians appear in a different color than vehicles.
|
||||
|
||||
The server provides an image with the tag information encoded in the red
|
||||
channel. A pixel with a red value of x displays an object with tag x. The
|
||||
The server provides an image with the tag information **encoded in the red
|
||||
channel**. A pixel with a red value of x displays an object with tag x. The
|
||||
following tags are currently available
|
||||
|
||||
Value | Tag
|
||||
|
|
@ -106,6 +112,13 @@ Value | Tag
|
|||
12 | TrafficSigns
|
||||
|
||||
This is implemented by tagging every object in the scene before hand (either at
|
||||
begin play or on spawn). The objects are classified by their relative file
|
||||
system path in the project. E.g., every mesh stored in the "pedestrians" folder
|
||||
it's tagged as pedestrian.
|
||||
begin play or on spawn). The objects are classified by their relative file path
|
||||
in the project. E.g., every mesh stored in the
|
||||
_"Unreal/CarlaUE4/Content/Static/Pedestrians"_ folder it's tagged as pedestrian.
|
||||
|
||||
!!! note
|
||||
**Adding new tags**:
|
||||
At the moment adding new tags is not very flexible and requires to modify
|
||||
the C++ code. Add a new label to the `ECityObjectLabel` enum in "Tagger.h",
|
||||
and its corresponding filepath check inside `GetLabelByFolderName()`
|
||||
function in "Tagger.cpp".
|
||||
|
|
|
|||
|
|
@ -12,6 +12,10 @@ CARLA is composed by the following modules
|
|||
- Carla plugin for Unreal Engine: "Unreal/CarlaUE4/Plugins/Carla"
|
||||
- CarlaServer: "Util/CarlaServer"
|
||||
|
||||
!!! tip
|
||||
Documentation for the C++ code can be generated by running
|
||||
[Doxygen](http://www.doxygen.org) in the main folder of CARLA project.
|
||||
|
||||
Python client API
|
||||
-----------------
|
||||
|
||||
|
|
|
|||
|
|
@ -0,0 +1,89 @@
|
|||
Running CARLA without Display and Selecting GPUs
|
||||
------
|
||||
|
||||
|
||||
This tutorial is designed for:
|
||||
- Remote server users that have several nvidia graphical cards and want to effectively use CARLA on all GPUs.
|
||||
- Desktop users who want to use the GPU that is not plugged on the screen for rendering CARLA.
|
||||
|
||||
|
||||
On this tutorial you will learn.
|
||||
- How to configure your server to have nvidia working on rendering without a display attached.
|
||||
- How to use VNC + VGL to simulate a display connected to any GPU you have in your machine.
|
||||
- And Finally, how to run CARLA in this environment
|
||||
|
||||
This tutorial was tested in Ubuntu 16.04 and using NVIDIA 384.11 drivers.
|
||||
## Preliminaries
|
||||
|
||||
A few things need to be working in your server before.
|
||||
Latest NVIDIA Drivers, OpenGL, VirtualGL(VGL), TurboVNC 2.11, ,
|
||||
#### NVIDIA Drivers
|
||||
Download and install NVIDIA-drivers with typical tutorials
|
||||
http://www.nvidia.es/Download/index.aspx
|
||||
#### OpenGL
|
||||
Openg GL is necessary for Virtual GL. Normally OpenGL
|
||||
can be installed through apt.
|
||||
|
||||
sudo apt-get install freeglut3-dev mesa-utils
|
||||
#### VGL
|
||||
Follow this tutorial and install vgl: <br>
|
||||
[Installing VGL](https://virtualgl.org/vgldoc/2_2_1/#hd004001)
|
||||
|
||||
#### TurboVNC
|
||||
Follow the tutorial below to install TurboVNC 2.11:<br>
|
||||
[Installing TurboVNC](https://cdn.rawgit.com/TurboVNC/turbovnc/2.1.1/doc/index.html#hd005001)
|
||||
|
||||
WARNING: Take care on which VNC you install as it may not be compatible with Unreal. The one above was the only one that worked for me.
|
||||
#### Extra Packages
|
||||
These extra packages were necessary to make unreal to work.
|
||||
|
||||
sudo apt install x11-xserver-utils libxrandr-dev
|
||||
|
||||
|
||||
|
||||
|
||||
#### Configure your X
|
||||
|
||||
You must generate a X compatible with your nvdia and compatible to run without display. For that, the following command worked:
|
||||
|
||||
sudo nvidia-xconfig -a --use-display-device=None --virtual=1280x1024
|
||||
|
||||
|
||||
|
||||
## Emulating The Virtual Display
|
||||
|
||||
|
||||
Run your own Xorg. Here I use number 7, but it could be labeled with any free number.
|
||||
|
||||
sudo nohup Xorg :7 &
|
||||
|
||||
Run an auxiliary remote VNC-Xserver. This will create a
|
||||
virtual display "8".
|
||||
|
||||
/opt/TurboVNC/bin/vncserver :8
|
||||
|
||||
If everything is working fine the following command
|
||||
should run smoothly.
|
||||
|
||||
DISPLAY=:8 vglrun -d :7.0 glxinfo
|
||||
Note. This will run glxinfo on Xserver 7, device 0. This means you are selecting the GPU 0 on your machine. To run on other GPU, such as GPU 1 run:
|
||||
|
||||
DISPLAY=:8 vglrun -d :7.1 glxinfo
|
||||
|
||||
#### Extra
|
||||
|
||||
If you want disable the need of sudo when creating the 'nohup Xorg'
|
||||
go to the '/etc/X11/Xwrapper.config' file and change 'allowed_users=console' to 'allowed_users=anybody'
|
||||
|
||||
It may be needed to stop all Xorg servers before running nohup Xorg.
|
||||
The command for that could change depending on your system. Generally for Ubuntu 16.04
|
||||
you should use:
|
||||
|
||||
sudo service lightdm stop
|
||||
|
||||
|
||||
## Running CARLA
|
||||
|
||||
Now, finally, to run CARLA on a certain gpu_number placed in a certain $CARLA_PATH, run.
|
||||
|
||||
DISPLAY=:8 vglrun -d :7.<gpu_number> $CARLA_PATH/CarlaUE4/Binaries/Linux/CarlaUE4
|
||||
|
|
@ -0,0 +1,25 @@
|
|||
Coding standard
|
||||
===============
|
||||
|
||||
> _This document is a work in progress and might be incomplete._
|
||||
|
||||
General
|
||||
-------
|
||||
|
||||
* Use spaces, not tabs.
|
||||
* Avoid adding trailing whitespace as it creates noise in the diffs.
|
||||
* Comments should not exceed 80 columns, code may exceed this limit a bit in rare occasions if it results in clearer code.
|
||||
|
||||
Python
|
||||
------
|
||||
|
||||
* All code must be compatible with Python 2.7, 3.5, and 3.6.
|
||||
* [Pylint](https://www.pylint.org/) should not give any error or warning (few exceptions apply with external classes like `numpy`, see our `.pylintrc`).
|
||||
* Python code follows [PEP8 style guide](https://www.python.org/dev/peps/pep-0008/) (use `autopep8` whenever possible).
|
||||
|
||||
C++
|
||||
---
|
||||
|
||||
* Compilation should not give any error or warning (`clang++ -Wall -Wextra -std=C++14`).
|
||||
* Unreal C++ code (CarlaUE4 and Carla plugin) follow the [Unreal Engine's Coding Standard](https://docs.unrealengine.com/latest/INT/Programming/Development/CodingStandard/) with the exception of using spaces instead of tabs.
|
||||
* CarlaServer uses [Google's style guide](https://google.github.io/styleguide/cppguide.html).
|
||||
196
Docs/faq.md
196
Docs/faq.md
|
|
@ -1,7 +1,20 @@
|
|||
CARLA F.A.Q.
|
||||
============
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
What is the recommended hardware to run CARLA?
|
||||
</h4></summary>
|
||||
|
||||
#### What is the expected disk space needed for building CARLA?
|
||||
CARLA is a very performance demanding software, at the very minimum you would
|
||||
need a computer with a dedicated GPU capable of running Unreal Engine. See
|
||||
[Unreal Engine's recommended hardware](https://wiki.unrealengine.com/Recommended_Hardware).
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
What is the expected disk space needed for building CARLA?
|
||||
</h4></summary>
|
||||
|
||||
Building CARLA from source requires about 15GB of disk space, not counting
|
||||
Unreal Engine installation.
|
||||
|
|
@ -10,7 +23,109 @@ However, you will also need to build and install Unreal Engine, which on Linux
|
|||
requires much more disk space as it keeps all the intermediate files,
|
||||
[see this thread](https://answers.unrealengine.com/questions/430541/linux-engine-size.html).
|
||||
|
||||
#### Is it possible to dump images from the CARLA server view?
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
I downloaded CARLA source from GitHub, where is the "CarlaUE4.sh" script?
|
||||
</h4></summary>
|
||||
|
||||
There is no "CarlaUE4.sh" script in the source version of CARLA, you need to
|
||||
follow the instructions in the [documentation](http://carla.readthedocs.io) for
|
||||
building CARLA from source.
|
||||
|
||||
Once you open the project in the Unreal Editor, you can hit Play to test CARLA.
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Setup.sh fails to download content, can I skip this step?
|
||||
</h4></summary>
|
||||
|
||||
It is possible to skip the download step by passing the `-s` argument to the
|
||||
setup script
|
||||
|
||||
$ ./Setup.sh -s
|
||||
|
||||
Bear in mind that if you do so, you are supposed to manually download and
|
||||
extract the content package yourself, check out the last output of the Setup.sh
|
||||
for instructions or run
|
||||
|
||||
$ ./Update.sh -s
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Can I run the server from within Unreal Editor?
|
||||
</h4></summary>
|
||||
|
||||
Yes, you can connect the Python client to a server running within Unreal Editor
|
||||
as if it was the standalone server.
|
||||
|
||||
Go to **"Unreal/CarlaUE4/Config/CarlaSettings.ini"** (this file should have been
|
||||
created by the Setup.sh) and enable networking. If for whatever reason you don't
|
||||
have this file, just create it and add the following
|
||||
|
||||
```ini
|
||||
[CARLA/Server]
|
||||
UseNetworking=true
|
||||
```
|
||||
|
||||
Now when you hit Play the editor will hang until a client connects.
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Why Unreal Editor hangs after hitting Play?
|
||||
</h4></summary>
|
||||
|
||||
This is most probably happening because CARLA is starting in server mode. Check
|
||||
your **"Unreal/CarlaUE4/Config/CarlaSettings.ini"** and set
|
||||
|
||||
```ini
|
||||
[CARLA/Server]
|
||||
UseNetworking=false
|
||||
```
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
How can I create a binary version of CARLA?
|
||||
</h4></summary>
|
||||
|
||||
To compile a binary (packaged) version of CARLA, open the CarlaUE4 project with
|
||||
Unreal Editor, go to the menu "File -> Package Project", and select your
|
||||
platform. This takes a while, but in the end it should generate a packaged
|
||||
version of CARLA to execute without Unreal Editor.
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Why do I have very low FPS when running the server in Unreal Editor?
|
||||
</h4></summary>
|
||||
|
||||
UE4 Editor goes to a low performance mode when out of focus. It can be disabled
|
||||
in the editor preferences. Go to "Edit->Editor Preferences->Performance" and
|
||||
disable the "Use Less CPU When in Background" option.
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Is it possible to dump images from the CARLA server view?
|
||||
</h4></summary>
|
||||
|
||||
Yes, this is an Unreal Engine feature. You can dump the images of the server
|
||||
camera by running CARLA with
|
||||
|
|
@ -19,17 +134,70 @@ camera by running CARLA with
|
|||
|
||||
Images are saved to "CarlaUE4/Saved/Screenshots/LinuxNoEditor".
|
||||
|
||||
#### I downloaded CARLA source from GitHub, where is the "CarlaUE4.sh" script?
|
||||
</details>
|
||||
|
||||
There is no "CarlaUE4.sh" script in the source version of CARLA, you need to
|
||||
follow the instructions in the [documentation](http://carla.readthedocs.io) on
|
||||
building CARLA from source.
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Fatal error: 'version.h' has been modified since the precompiled header.
|
||||
</h4></summary>
|
||||
|
||||
Once you open the project in the Unreal Editor, you can hit Play to test CARLA.
|
||||
This happens from time to time due to Linux updates. It is possible to force a
|
||||
rebuild of all the project files with
|
||||
|
||||
#### How can I create a binary version of CARLA?
|
||||
$ cd Unreal/CarlaUE4/
|
||||
$ make CarlaUE4Editor ARGS=-clean
|
||||
$ make CarlaUE4Editor
|
||||
|
||||
To compile a binary (packaged) version of CARLA, open the CarlaUE4 project with
|
||||
Unreal Editor, go to the menu “File -> Package Project”, and select your
|
||||
platform. This takes a while, but in the end it should generate a packaged
|
||||
version of CARLA to execute without Unreal Editor.
|
||||
It takes a long time but fixes the issue. Sometimes a reboot is also needed.
|
||||
|
||||
</details>
|
||||
|
||||
<!-- ======================================================================= -->
|
||||
<details>
|
||||
<summary><h5 style="display:inline">
|
||||
Fatal error: 'carla/carla_server.h' file not found.
|
||||
</h4></summary>
|
||||
|
||||
This indicates that the CarlaServer dependency failed to compile.
|
||||
|
||||
Please follow the instructions at
|
||||
[How to build on Linux](http://carla.readthedocs.io/en/latest/how_to_build_on_linux/).
|
||||
|
||||
Make sure that the Setup script does print _"Success!"_ at the end
|
||||
|
||||
$ ./Setup.sh
|
||||
...
|
||||
...
|
||||
****************
|
||||
*** Success! ***
|
||||
****************
|
||||
|
||||
Then check if CarlaServer compiles without errors running make
|
||||
|
||||
$ make
|
||||
|
||||
It should end printing something like
|
||||
|
||||
```
|
||||
[1/1] Install the project...
|
||||
-- Install configuration: "Release"
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++abi.so.1
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++abi.so.1.0
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++.so.1
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++.so.1.0
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++.so
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/shared/libc++abi.so
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/lib/libc++abi.a
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/lib/libboost_system.a
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/lib/libprotobuf.a
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/include/carla
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/include/carla/carla_server.h
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/lib/libcarlaserver.a
|
||||
-- Installing: Unreal/CarlaUE4/Plugins/Carla/CarlaServer/bin/test_carlaserver
|
||||
-- Set runtime path of "Unreal/CarlaUE4/Plugins/Carla/CarlaServer/bin/test_carlaserver" to ""
|
||||
```
|
||||
|
||||
If so you can safely run Rebuild.sh.
|
||||
|
||||
</details>
|
||||
|
|
|
|||
|
|
@ -20,13 +20,15 @@ change your default clang version to compile Unreal
|
|||
Build Unreal Engine
|
||||
-------------------
|
||||
|
||||
!!! note
|
||||
Unreal Engine repositories are set to private. In order to gain access you
|
||||
need to add your GitHub username when you sign up at
|
||||
[www.unrealengine.com](https://www.unrealengine.com).
|
||||
|
||||
Download and compile Unreal Engine 4.17. Here we will assume you install it at
|
||||
"~/UnrealEngine_4.17", but you can install it anywhere, just replace the path
|
||||
where necessary.
|
||||
|
||||
Unreal Engine repositories are set to private. In order to gain access you need
|
||||
to add your GitHub username when you sign up at https://unrealengine.com.
|
||||
|
||||
$ git clone --depth=1 -b 4.17 https://github.com/EpicGames/UnrealEngine.git ~/UnrealEngine_4.17
|
||||
$ cd ~/UnrealEngine_4.17
|
||||
$ ./Setup.sh && ./GenerateProjectFiles.sh && make
|
||||
|
|
|
|||
|
|
@ -38,15 +38,9 @@ The "carla" Python module provides a basic API for communicating with the CARLA
|
|||
server. In the "PythonClient" folder we provide a couple of examples on how to
|
||||
use this API. We recommend Python 3, but they are also compatible with Python 2.
|
||||
|
||||
The basic functionality requires only the protobuf module to be installed
|
||||
Install the dependencies with
|
||||
|
||||
$ sudo apt-get install python3 python3-pip
|
||||
$ sudo pip3 install protobuf
|
||||
|
||||
However, other operations as handling images require some extra modules, and the
|
||||
"manual_control.py" example requires pygame
|
||||
|
||||
$ sudo pip3 install numpy Pillow pygame
|
||||
$ pip install -r PythonClient/requirements.txt
|
||||
|
||||
The script "PythonClient/client_example.py" provides basic functionality for
|
||||
controlling the vehicle and saving images to disk. Run the help command to see
|
||||
|
|
|
|||
Binary file not shown.
|
After Width: | Height: | Size: 1.5 MiB |
Binary file not shown.
|
After Width: | Height: | Size: 93 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 128 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 7.3 KiB |
|
|
@ -7,8 +7,9 @@ CARLA Documentation
|
|||
* [CARLA settings](carla_settings.md)
|
||||
* [Measurements](measurements.md)
|
||||
* [Cameras and sensors](cameras_and_sensors.md)
|
||||
* [CARLA without Display and Selecting GPUs](carla_headless.md)
|
||||
* [Benchmark](benchmark.md)
|
||||
* [F.A.Q.](faq.md)
|
||||
* [Troubleshooting](troubleshooting.md)
|
||||
|
||||
#### Building from source
|
||||
|
||||
|
|
@ -19,6 +20,8 @@ CARLA Documentation
|
|||
#### Contributing
|
||||
|
||||
* [Contribution guidelines](CONTRIBUTING.md)
|
||||
* [Coding standard](coding_standard.md)
|
||||
* [Code of conduct](CODE_OF_CONDUCT.md)
|
||||
|
||||
#### Development
|
||||
|
||||
|
|
|
|||
|
|
@ -0,0 +1,12 @@
|
|||
<!--
|
||||
|
||||
Thanks for contributing to CARLA!
|
||||
|
||||
If you are asking a question please make sure your question was not asked before
|
||||
by searching among the existing issues. Also make sure you have read our
|
||||
documentation and FAQ at carla.readthedocs.io.
|
||||
|
||||
If you are reporting an issue, please provide the CARLA version number you are
|
||||
using and your Platform/OS.
|
||||
|
||||
-->
|
||||
|
|
@ -1,10 +1,50 @@
|
|||
# Map customization
|
||||
|
||||
## Creating a new map
|
||||
|
||||
!!! Bug
|
||||
Creating a map from scratch with the Carla tools causes a crash with
|
||||
UE4.17.2 ([Issue #99](https://github.com/carla-simulator/carla/issues/99)),
|
||||
this guide will suggest duplicating an existing level instead of creating
|
||||
one from scratch.
|
||||
|
||||
### Requirements
|
||||
|
||||
- Checkout and build Carla from source on [Linux](how_to_build_on_linux.md) or [Windows](how_to_build_on_windows.md)
|
||||
|
||||
### Creating
|
||||
|
||||
- Duplicate an existing map
|
||||
- Remove everything you don't need from the map
|
||||
- Keep the folder "Lighting", "AtmosphericFog", "PostProcessVol" and "CarlaMapGenerator" this will keep the climate working as intended and the post process saved.
|
||||
- It might be interesting to keep the empty level as a template and duplicate it before starting to populate it.
|
||||
- In the CarlaMapGenerator, there is a field "seed". You can change the map by altering that seed and clicking "Trigger Road Map Generation". "Save Road Map To Disk" should also be checked.
|
||||
- You can change the seed until you have a map you are satisfied with.
|
||||
- After that you can place new PlayerStarts at the places you want the cars to be spawned.
|
||||
- The AI already works, but the cars won't act randomly. Vehicles will follow the instructions given by the RoadMapGenerator. They will follow the road easily while in straight roads but wont so much when entering Intersections:
|
||||
|
||||
> (This is a debug view of the instructions the road gives to the Vehicle. They will always follow the green arrows, the white points are shared points between one or more routes, by default they order the vehicle to continue straight; Black points are off the road, the vehicle gets no instructions and drives to the left, trying to get back to the road)
|
||||
|
||||
- To get a random behavior, you have to place IntersectionEntrances, this will let you redefine the direction the vehicle will take overwriting the directions given by the road map (until they finish their given order).
|
||||
(See the two example towns how it exactly works).
|
||||
- Before version 0.7.1: For every entrance you'll have to create a series of empty actors that will be the waypoints to guide the car through the intersection; Then you'll have to assign the corresponding actors to every Path
|
||||
- After version 0.7.1: Every IntersectionEntrance has an array called routes, adding an element to this creates an editable spline in the world with the first point on the IntersectionEntrance (You might have to select another object before you can see it) This spline defines the possible routes any car will take when entering the intersection (as the Empty actors did before) you might configure this routes as you would edit any Unreal spline. Each route will create an element in the field bellow: "Probabilities" every number in this array defines the chances of any vehicle to take the corresponding route.
|
||||
- To change the speed of the car, use the SpeedLimiters. They are straightforward to use. (Make sure you limit the speed for the corners, otherwise the cars will try and fail to take them at full speed)
|
||||
- Traffic lights need to be scripted to avoid traffic accidents.
|
||||
Every street at a crossing should have its own turn at green without the other streets having green.
|
||||
- Then you can populate the world with landscape and buildings.
|
||||
|
||||
## Blueprint Assets
|
||||
|
||||
This are the specific blueprint assets created to help building the environment.
|
||||
|
||||
## MultipleFloorBuilding:
|
||||
|
||||
The purpose of this blueprint is to make repeating and varying tall buildings a bit easier. Provided a Base, a MiddleFloor and a roof; this blueprint repeats the middle floor to the desired number of stores and tops it whith the last floor given some conditions:
|
||||
The purpose of this blueprint is to make repeating and varying tall buildings a
|
||||
bit easier. Provided a Base, a MiddleFloor and a roof; this blueprint repeats
|
||||
the middle floor to the desired number of stores and tops it with the last floor
|
||||
given some conditions:
|
||||
|
||||
- All model pivots should be in the bottom center of the Specific mesh.
|
||||
- Al models must start and end exactly where the repetition happen.
|
||||
|
||||
|
|
@ -14,53 +54,78 @@ This blueprint is controlled by this 6 specific Parameters:
|
|||
- Floor: The mesh to be repeated along the building.
|
||||
- Roof: Final mesh to top the building.
|
||||
- FloorNumber: Number of stores of the building.
|
||||
- FloorHeightOffset: Adjust The placement of every floor verticaly.
|
||||
- RoofOffset: Adjust the placement of the roof verticaly.
|
||||
- FloorHeightOffset: Adjust The placement of every floor vertically.
|
||||
- RoofOffset: Adjust the placement of the roof vertically.
|
||||
|
||||
All of This parameters can be modified once this blueprint is placed in the world.
|
||||
All of This parameters can be modified once this blueprint is placed in the
|
||||
world.
|
||||
|
||||
## SplinemeshRepeater:
|
||||
|
||||
!!! Bug
|
||||
See [#35 SplineMeshRepeater loses its collider mesh](https://github.com/carla-simulator/carla/issues/35)
|
||||
|
||||
|
||||
### Standard use:
|
||||
|
||||
SplineMeshRepeater "Content/Blueprints/SplineMeshRepeater" is a tool included in the Carla Project to help building urban environments; It repeats and aligns a specific choosen mesh along a [Spline](https://docs.unrealengine.com/latest/INT/Engine/BlueprintSplines/Reference/SplineEditorTool/index.html) (Unreal Component). Its principal function is to build Tipicaly tiled and repetitive structures as Walls, Roads, Bridges, Fences... Once the actor is placed into the world the spline can be modified so the object gets the desired form. Each Point Defining the spline Generates a new tile so that as more points the Spline has, the more defined it will be, but also heavyer on the world. This actor is defined by the following parameters:
|
||||
|
||||
SplineMeshRepeater "Content/Blueprints/SplineMeshRepeater" is a tool included in
|
||||
the Carla Project to help building urban environments; It repeats and aligns a
|
||||
specific chosen mesh along a
|
||||
[Spline](https://docs.unrealengine.com/latest/INT/Engine/BlueprintSplines/Reference/SplineEditorTool/index.html)
|
||||
(Unreal Component). Its principal function is to build Typically tiled and
|
||||
repetitive structures as Walls, Roads, Bridges, Fences... Once the actor is
|
||||
placed into the world the spline can be modified so the object gets the desired
|
||||
form. Each Point Defining the spline Generates a new tile so that as more points
|
||||
the Spline has, the more defined it will be, but also heavier on the world. This
|
||||
actor is defined by the following parameters:
|
||||
|
||||
- StaticMesh: The mesh to be repeated along the spline.
|
||||
- ForWardAxis: Changes the mesh axis to be alingned with the spline.
|
||||
- ForWardAxis: Changes the mesh axis to be aligned with the spline.
|
||||
- Material: Overrides the mesh' default material.
|
||||
- Colission Enabled: Chooses the tipe of colission to use.
|
||||
- Collision Enabled: Chooses the type of collision to use.
|
||||
- Gap distance: Places a Gap between each repeated mesh, for repetitive non continuous walls: bush chains, bollards...
|
||||
|
||||
|
||||
(Last three variables are specific for some particular assets to be defined in the next point) A requisite to create assets compatibles with this componenis is that all the meshes have their pivot placed wherever the repetition starts in the lower point possible with the rest of the mesh pointing positive (Preferably by the X axis)
|
||||
(Last three variables are specific for some particular assets to be defined in
|
||||
the next point) A requisite to create assets compatibles with this component is
|
||||
that all the meshes have their pivot placed wherever the repetition starts in
|
||||
the lower point possible with the rest of the mesh pointing positive (Preferably
|
||||
by the X axis)
|
||||
|
||||
|
||||
### Specific Walls (Dynamic material)
|
||||
|
||||
In the project folder "Content/Static/Walls" are included some specific assets to be used with this SplineMeshRepeater with a series of special caracteristics. The uv space of this meshes and their materials are the same for all of them, making them excangeable. each material is composed of three different surfaces the last three parameters slightly modify the color of this surfaces:
|
||||
In the project folder "Content/Static/Walls" are included some specific assets
|
||||
to be used with this SplineMeshRepeater with a series of special
|
||||
characteristics. The UV space of this meshes and their materials are the same
|
||||
for all of them, making them exchangeable. each material is composed of three
|
||||
different surfaces the last three parameters slightly modify the color of this
|
||||
surfaces:
|
||||
|
||||
- MainMaterialColor: Change the main material of the Wall
|
||||
- DetailsColor: Change the color of the details (if any)
|
||||
- TopWallColor: Cambia el color de la cubierta del muro (if any)
|
||||
- TopWallColor: Change the color of the wall cover (if any)
|
||||
|
||||
To add elements that profit from this functions exist in the (Carpeta) folder the GardenWallMask File that defines the uv space to place the materials: (Blue space: MainMaterial; green space: Details; red space TopWall).
|
||||
|
||||
Between the material masters is WallMaster which is going to be the master of the materials using this function. An instance of this material will be created and the correspondent textures will be added. This material includes the following parameters to be modified by the material to use:
|
||||
|
||||
- Normal Flattener: Slightly modifies the normal map values to exagerate it or flatten it.
|
||||
- RoughnessCorrection: Para corregir El valor de rugosidad dado por la textura.
|
||||
|
||||
The rest of the parameters are the mask the textures and the color corrections that won't be modified in this instance but in the blueprint that will be launched into the world.
|
||||
To add elements that profit from this functions exist the GardenWallMask File that defines the uv space to place the materials: (Blue space: MainMaterial; green space: Details; red space TopWall).
|
||||
|
||||
Between the material masters is WallMaster which is going to be the master of
|
||||
the materials using this function. An instance of this material will be created
|
||||
and the correspondent textures will be added. This material includes the
|
||||
following parameters to be modified by the material to use:
|
||||
|
||||
- Normal Flattener: Slightly modifies the normal map values to exaggerate it or flatten it.
|
||||
- RoughnessCorrection: Changes the Roughness value given by the texture.
|
||||
|
||||
The rest of the parameters are the mask the textures and the color corrections
|
||||
that won't be modified in this instance but in the blueprint that will be
|
||||
launched into the world.
|
||||
|
||||
## Weather
|
||||
This is the actor in charge of modifying all the lighting, environmental actors an anything that affects the impression of the climate. It runs automaticaly with the game when is not specified otherwise In the Config.Ini but has Its own actor to launch in editor mode to configure the climatic conditions. To fuly work It will need One of each of the folowing actors: SkySphere, Skylight, Postprocess Volume (Boundless) And Light Source to exist in the world.
|
||||
|
||||
This is the actor in charge of modifying all the lighting, environmental actors
|
||||
an anything that affects the impression of the climate. It runs automatically
|
||||
with the game when is not specified otherwise In the Config. Ini but has Its own
|
||||
actor to launch in editor mode to configure the climatic conditions. To fully
|
||||
work It will need One of each of the following actors: SkySphere, Skylight,
|
||||
Postprocess Volume (Boundless) And Light Source to exist in the world.
|
||||
|
||||
- SunPolarAngle: polar angle of the sun, determines time of the day
|
||||
- SunAzimuthAngle: adds to the location of the sun in the current level
|
||||
|
|
@ -68,7 +133,7 @@ This is the actor in charge of modifying all the lighting, environmental actors
|
|||
- SunDirectionalLightIntensity: Intensity of the sunlight
|
||||
- SunDirectionalLightColor: Color of the sunlight
|
||||
- SunIndirectLightIntensity: intensity of the bounces of the main light
|
||||
- CloudOpacity: visivility of the cloud rendering on the skybox
|
||||
- CloudOpacity: visibility of the cloud rendering on the skybox
|
||||
- HorizontFalloff: determines the height of the gradient between the zenith and horizon color
|
||||
- ZenithColor: Defines the color of the zenith.
|
||||
- HorizonColor: Defines the color of the horizon.
|
||||
|
|
@ -79,7 +144,7 @@ This is the actor in charge of modifying all the lighting, environmental actors
|
|||
- Precipitation: Defines if any precipitation is active.
|
||||
- PrecipitationType: the type of precipitation to active.
|
||||
- PrecipitationAmount: the quantity of the chosen precipitation.
|
||||
- PrecipitationAccumulation: the acumulation of the chosen precipitation.
|
||||
- PrecipitationAccumulation: the accumulation of the chosen precipitation.
|
||||
- bWind: defines if there is any wind.
|
||||
- WindIntensity: defines the wind intensity.
|
||||
- WindAngle: defines the wind direction.
|
||||
|
|
@ -89,5 +154,7 @@ This is the actor in charge of modifying all the lighting, environmental actors
|
|||
- CameraPostProcessParameters.AutoExposureMaxBrightness: defines the maximum brightness the autoexposure will count as right in the final image.
|
||||
- CameraPostProcessParameters.AutoExposureBias: Darkens or brightens the final image towards a defined bias.
|
||||
|
||||
|
||||
You can have as many different configurations saved in the proyect as you want and choose the configuration to aply while on the build, through the [settings file](carla_settings.md); or in the editor while building the level or testing.
|
||||
You can have as many different configurations saved in the project as you want
|
||||
and choose the configuration to apply while on the build, through the
|
||||
[settings file](carla_settings.md); or in the editor while building the level or
|
||||
testing.
|
||||
|
|
|
|||
|
|
@ -0,0 +1,32 @@
|
|||
<!--
|
||||
|
||||
Thanks for sending a pull request! Please make sure you click the link above to
|
||||
view the contribution guidelines, then fill out the blanks below.
|
||||
|
||||
Checklist:
|
||||
|
||||
- [ ] Your branch is up-to-date with the `master` branch and tested with latest changes
|
||||
- [ ] Extended the README / documentation, if necessary
|
||||
- [ ] Code compiles correctly
|
||||
- [ ] All tests passing
|
||||
- [ ] `make check`
|
||||
- [ ] `pylint --disable=R,C --rcfile=PythonClient/.pylintrc PythonClient/carla PythonClient/*.py`
|
||||
- [ ] `cppcheck . -iBuild -i.pb.cc --enable=warning`
|
||||
|
||||
-->
|
||||
|
||||
#### Description
|
||||
|
||||
<!-- Please explain the changes you made here as detailed as possible. -->
|
||||
|
||||
Fixes # <!-- If fixes an issue, please add here the issue number. -->
|
||||
|
||||
#### Where has this been tested?
|
||||
|
||||
* **Platform(s):** ...
|
||||
* **Python version(s):** ...
|
||||
* **Unreal Engine version(s):** ...
|
||||
|
||||
#### Possible Drawbacks
|
||||
|
||||
<!-- What are the possible side-effects or negative impacts of the code change? -->
|
||||
|
|
@ -1,28 +0,0 @@
|
|||
Troubleshooting
|
||||
===============
|
||||
|
||||
#### Editor hangs after hitting Play
|
||||
|
||||
This is most probably happening because CARLA is started in server mode. Check
|
||||
in your CarlaSettings.ini file ("./Unreal/CarlaUE4/Config/CarlaSettings.ini")
|
||||
and set
|
||||
|
||||
```ini
|
||||
[CARLA/Server]
|
||||
UseNetworking=false
|
||||
```
|
||||
|
||||
#### Very low FPS in editor when not in focus
|
||||
|
||||
UE4 Editor goes to a low performance mode when out of focus. It can be disabled
|
||||
in the editor preferences. Go to "Edit->Editor Preferences->Performance" and
|
||||
disable the "Use Less CPU When in Background" option.
|
||||
|
||||
#### Fatal error: file '/usr/include/linux/version.h' has been modified since the precompiled header
|
||||
|
||||
This happens from time to time due to Linux updates. It is possible to force a
|
||||
rebuild of all the project files with
|
||||
|
||||
$ cd Unreal/CarlaUE4/
|
||||
$ make CarlaUE4Editor ARGS=-clean
|
||||
$ make CarlaUE4Editor
|
||||
2
LICENSE
2
LICENSE
|
|
@ -1,7 +1,7 @@
|
|||
MIT License
|
||||
|
||||
Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
Barcelona (UAB).
|
||||
|
||||
Permission is hereby granted, free of charge, to any person obtaining a copy
|
||||
of this software and associated documentation files (the "Software"), to deal
|
||||
|
|
|
|||
|
|
@ -0,0 +1,176 @@
|
|||
#! /bin/bash
|
||||
|
||||
################################################################################
|
||||
# Packages a CARLA build.
|
||||
################################################################################
|
||||
|
||||
set -e
|
||||
|
||||
DOC_STRING="Makes a packaged version of CARLA for distribution.
|
||||
Please make sure to run Rebuild.sh before!"
|
||||
|
||||
USAGE_STRING="Usage: $0 [-h|--help] [--no-packaging] [--no-zip] [--clean-intermediate]"
|
||||
|
||||
# ==============================================================================
|
||||
# -- Parse arguments -----------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
DO_PACKAGE=true
|
||||
DO_COPY_FILES=true
|
||||
DO_TARBALL=true
|
||||
DO_CLEAN_INTERMEDIATE=false
|
||||
|
||||
OPTS=`getopt -o h --long help,no-packaging,no-zip,clean-intermediate -n 'parse-options' -- "$@"`
|
||||
|
||||
if [ $? != 0 ] ; then echo "$USAGE_STRING" ; exit 2 ; fi
|
||||
|
||||
eval set -- "$OPTS"
|
||||
|
||||
while true; do
|
||||
case "$1" in
|
||||
--no-packaging )
|
||||
DO_PACKAGE=false
|
||||
shift ;;
|
||||
--no-zip )
|
||||
DO_TARBALL=false
|
||||
shift ;;
|
||||
--clean-intermediate )
|
||||
DO_CLEAN_INTERMEDIATE=true
|
||||
shift ;;
|
||||
-h | --help )
|
||||
echo "$DOC_STRING"
|
||||
echo "$USAGE_STRING"
|
||||
exit 1
|
||||
;;
|
||||
* )
|
||||
break ;;
|
||||
esac
|
||||
done
|
||||
|
||||
# ==============================================================================
|
||||
# -- Set up environment --------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
|
||||
pushd "$SCRIPT_DIR" >/dev/null
|
||||
|
||||
REPOSITORY_TAG=`git describe --tags --dirty --always`
|
||||
|
||||
echo "Packaging version '$REPOSITORY_TAG'."
|
||||
|
||||
UNREAL_PROJECT_FOLDER=${PWD}/Unreal/CarlaUE4
|
||||
DIST_FOLDER=${PWD}/Dist
|
||||
BUILD_FOLDER=${DIST_FOLDER}/${REPOSITORY_TAG}
|
||||
|
||||
function fatal_error {
|
||||
echo -e "\033[0;31mERROR: $1\033[0m"
|
||||
exit 1
|
||||
}
|
||||
|
||||
function log {
|
||||
echo -e "\033[0;33m$1\033[0m"
|
||||
}
|
||||
|
||||
# ==============================================================================
|
||||
# -- Package project -----------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
if $DO_PACKAGE ; then
|
||||
|
||||
pushd "$UNREAL_PROJECT_FOLDER" >/dev/null
|
||||
|
||||
log "Packaging the project..."
|
||||
|
||||
if [ ! -d "${UE4_ROOT}" ]; then
|
||||
fatal_error "UE4_ROOT is not defined, or points to a non-existant directory, please set this environment variable."
|
||||
fi
|
||||
|
||||
rm -Rf ${BUILD_FOLDER}
|
||||
mkdir -p ${BUILD_FOLDER}
|
||||
|
||||
${UE4_ROOT}/Engine/Build/BatchFiles/RunUAT.sh BuildCookRun \
|
||||
-project="${PWD}/CarlaUE4.uproject" \
|
||||
-nocompileeditor -nop4 -cook -stage -archive -package \
|
||||
-clientconfig=Development -ue4exe=UE4Editor \
|
||||
-pak -prereqs -nodebuginfo \
|
||||
-targetplatform=Linux -build -CrashReporter -utf8output \
|
||||
-archivedirectory="${BUILD_FOLDER}"
|
||||
|
||||
popd >/dev/null
|
||||
|
||||
fi
|
||||
|
||||
if [[ ! -d ${BUILD_FOLDER}/LinuxNoEditor ]] ; then
|
||||
fatal_error "Failed to package the project!"
|
||||
fi
|
||||
|
||||
# ==============================================================================
|
||||
# -- Copy files (Python server, README, etc) -----------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
if $DO_COPY_FILES ; then
|
||||
|
||||
DESTINATION=${BUILD_FOLDER}/LinuxNoEditor
|
||||
|
||||
log "Copying extra files..."
|
||||
|
||||
cp -v ./LICENSE ${DESTINATION}/LICENSE
|
||||
cp -v ./CHANGELOG.md ${DESTINATION}/CHANGELOG
|
||||
cp -v ./Docs/release_readme.md ${DESTINATION}/README
|
||||
cp -v ./Docs/Example.CarlaSettings.ini ${DESTINATION}/Example.CarlaSettings.ini
|
||||
|
||||
rsync -vhr --delete --delete-excluded \
|
||||
--exclude "__pycache__" \
|
||||
--exclude "*.pyc" \
|
||||
--exclude ".*" \
|
||||
PythonClient/ ${DESTINATION}/PythonClient
|
||||
|
||||
echo
|
||||
|
||||
fi
|
||||
|
||||
# ==============================================================================
|
||||
# -- Zip the project -----------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
if $DO_TARBALL ; then
|
||||
|
||||
DESTINATION=${DIST_FOLDER}/CARLA_${REPOSITORY_TAG}.tar.gz
|
||||
SOURCE=${BUILD_FOLDER}/LinuxNoEditor
|
||||
|
||||
pushd "$SOURCE" >/dev/null
|
||||
|
||||
log "Packaging build..."
|
||||
|
||||
rm -f ./Manifest_NonUFSFiles_Linux.txt
|
||||
rm -Rf ./CarlaUE4/Saved
|
||||
rm -Rf ./Engine/Saved
|
||||
|
||||
tar -czvf ${DESTINATION} *
|
||||
|
||||
popd >/dev/null
|
||||
|
||||
fi
|
||||
|
||||
# ==============================================================================
|
||||
# -- Remove intermediate files -------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
if $DO_CLEAN_INTERMEDIATE ; then
|
||||
|
||||
log "Removing intermediate build..."
|
||||
|
||||
rm -Rf ${BUILD_FOLDER}
|
||||
|
||||
fi
|
||||
|
||||
# ==============================================================================
|
||||
# -- ...and we are done --------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
echo ""
|
||||
echo "****************"
|
||||
echo "*** Success! ***"
|
||||
echo "****************"
|
||||
|
||||
popd >/dev/null
|
||||
|
|
@ -0,0 +1,2 @@
|
|||
[pep8]
|
||||
max-line-length = 120
|
||||
|
|
@ -0,0 +1,2 @@
|
|||
include carla/planner/*.txt
|
||||
include carla/planner/*.png
|
||||
|
|
@ -0,0 +1,38 @@
|
|||
#!/usr/bin/env python2
|
||||
# -*- coding: utf-8 -*-
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
# @author: german,felipecode
|
||||
|
||||
|
||||
from __future__ import print_function
|
||||
import abc
|
||||
|
||||
from carla.planner.planner import Planner
|
||||
|
||||
|
||||
class Agent(object):
|
||||
def __init__(self, city_name):
|
||||
self.__metaclass__ = abc.ABCMeta
|
||||
self._planner = Planner(city_name)
|
||||
|
||||
def get_distance(self, start_point, end_point):
|
||||
path_distance = self._planner.get_shortest_path_distance(
|
||||
[start_point.location.x, start_point.location.y, 22]
|
||||
, [start_point.orientation.x, start_point.orientation.y, 22]
|
||||
, [end_point.location.x, end_point.location.y, 22]
|
||||
, [end_point.orientation.x, end_point.orientation.y, 22])
|
||||
# We calculate the timout based on the distance
|
||||
|
||||
return path_distance
|
||||
|
||||
@abc.abstractmethod
|
||||
def run_step(self, measurements, sensor_data, target):
|
||||
"""
|
||||
Function to be redefined by an agent.
|
||||
:param The measurements like speed, the image data and a target
|
||||
:returns A carla Control object, with the steering/gas/brake for the agent
|
||||
"""
|
||||
|
|
@ -0,0 +1,377 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
|
||||
import csv
|
||||
import datetime
|
||||
import math
|
||||
import os
|
||||
import abc
|
||||
import logging
|
||||
|
||||
|
||||
from builtins import input as input_data
|
||||
|
||||
|
||||
from carla.client import VehicleControl
|
||||
|
||||
def sldist(c1, c2):
|
||||
return math.sqrt((c2[0] - c1[0])**2 + (c2[1] - c1[1])**2)
|
||||
|
||||
|
||||
class Benchmark(object):
|
||||
|
||||
"""
|
||||
The Benchmark class, controls the execution of the benchmark by an
|
||||
Agent class.
|
||||
The benchmark class must be inherited
|
||||
"""
|
||||
def __init__(
|
||||
self,
|
||||
city_name,
|
||||
name_to_save,
|
||||
continue_experiment=False,
|
||||
save_images=False
|
||||
):
|
||||
|
||||
|
||||
self.__metaclass__ = abc.ABCMeta
|
||||
|
||||
self._city_name = city_name
|
||||
|
||||
|
||||
|
||||
self._base_name = name_to_save
|
||||
self._dict_stats = {'exp_id': -1,
|
||||
'rep': -1,
|
||||
'weather': -1,
|
||||
'start_point': -1,
|
||||
'end_point': -1,
|
||||
'result': -1,
|
||||
'initial_distance': -1,
|
||||
'final_distance': -1,
|
||||
'final_time': -1,
|
||||
'time_out': -1
|
||||
}
|
||||
|
||||
self._dict_rewards = {'exp_id': -1,
|
||||
'rep': -1,
|
||||
'weather': -1,
|
||||
'collision_gen': -1,
|
||||
'collision_ped': -1,
|
||||
'collision_car': -1,
|
||||
'lane_intersect': -1,
|
||||
'sidewalk_intersect': -1,
|
||||
'pos_x': -1,
|
||||
'pos_y': -1
|
||||
}
|
||||
|
||||
|
||||
self._experiments = self._build_experiments()
|
||||
# Create the log files and get the names
|
||||
self._suffix_name, self._full_name = self._create_log_record(name_to_save, self._experiments)
|
||||
# Get the line for the experiment to be continued
|
||||
self._line_on_file = self._continue_experiment(continue_experiment)
|
||||
|
||||
|
||||
|
||||
self._save_images = save_images
|
||||
self._image_filename_format = os.path.join(
|
||||
self._full_name, '_images/episode_{:s}/{:s}/image_{:0>5d}.jpg')
|
||||
|
||||
def run_navigation_episode(
|
||||
self,
|
||||
agent,
|
||||
carla,
|
||||
time_out,
|
||||
target,
|
||||
episode_name):
|
||||
|
||||
measurements, sensor_data = carla.read_data()
|
||||
carla.send_control(VehicleControl())
|
||||
|
||||
t0 = measurements.game_timestamp
|
||||
t1 = t0
|
||||
success = False
|
||||
measurement_vec = []
|
||||
frame = 0
|
||||
distance = 10000
|
||||
|
||||
while(t1 - t0) < (time_out * 1000) and not success:
|
||||
measurements, sensor_data = carla.read_data()
|
||||
|
||||
control = agent.run_step(measurements, sensor_data, target)
|
||||
|
||||
logging.info("Controller is Inputting:")
|
||||
logging.info('Steer = %f Throttle = %f Brake = %f ',
|
||||
control.steer, control.throttle, control.brake)
|
||||
|
||||
carla.send_control(control)
|
||||
|
||||
# measure distance to target
|
||||
if self._save_images:
|
||||
for name, image in sensor_data.items():
|
||||
image.save_to_disk(self._image_filename_format.format(
|
||||
episode_name, name, frame))
|
||||
|
||||
curr_x = measurements.player_measurements.transform.location.x
|
||||
curr_y = measurements.player_measurements.transform.location.y
|
||||
|
||||
measurement_vec.append(measurements.player_measurements)
|
||||
|
||||
t1 = measurements.game_timestamp
|
||||
|
||||
distance = sldist([curr_x, curr_y],
|
||||
[target.location.x, target.location.y])
|
||||
|
||||
logging.info('Status:')
|
||||
logging.info(
|
||||
'[d=%f] c_x = %f, c_y = %f ---> t_x = %f, t_y = %f',
|
||||
float(distance), curr_x, curr_y, target.location.x,
|
||||
target.location.y)
|
||||
|
||||
if distance < 200.0:
|
||||
success = True
|
||||
|
||||
frame += 1
|
||||
|
||||
if success:
|
||||
return 1, measurement_vec, float(t1 - t0) / 1000.0, distance
|
||||
return 0, measurement_vec, time_out, distance
|
||||
|
||||
def benchmark_agent(self, agent, carla):
|
||||
|
||||
if self._line_on_file == 0:
|
||||
# The fixed name considering all the experiments being run
|
||||
with open(os.path.join(self._full_name,
|
||||
self._suffix_name), 'w') as ofd:
|
||||
|
||||
w = csv.DictWriter(ofd, self._dict_stats.keys())
|
||||
w.writeheader()
|
||||
|
||||
with open(os.path.join(self._full_name,
|
||||
'details_' + self._suffix_name), 'w') as rfd:
|
||||
|
||||
rw = csv.DictWriter(rfd, self._dict_rewards.keys())
|
||||
rw.writeheader()
|
||||
start_task = 0
|
||||
start_pose = 0
|
||||
else:
|
||||
(start_task, start_pose) = self._get_pose_and_task(self._line_on_file)
|
||||
|
||||
logging.info(' START ')
|
||||
|
||||
for experiment in self._experiments[start_task:]:
|
||||
|
||||
positions = carla.load_settings(
|
||||
experiment.conditions).player_start_spots
|
||||
|
||||
for pose in experiment.poses[start_pose:]:
|
||||
for rep in range(experiment.repetitions):
|
||||
|
||||
start_point = pose[0]
|
||||
end_point = pose[1]
|
||||
|
||||
carla.start_episode(start_point)
|
||||
|
||||
logging.info('======== !!!! ==========')
|
||||
logging.info(' Start Position %d End Position %d ',
|
||||
start_point, end_point)
|
||||
|
||||
path_distance = agent.get_distance(
|
||||
positions[start_point], positions[end_point])
|
||||
euclidean_distance = \
|
||||
sldist([positions[start_point].location.x, positions[start_point].location.y],
|
||||
[positions[end_point].location.x, positions[end_point].location.y])
|
||||
|
||||
time_out = self._calculate_time_out(path_distance)
|
||||
# running the agent
|
||||
(result, reward_vec, final_time, remaining_distance) = \
|
||||
self.run_navigation_episode(
|
||||
agent, carla, time_out, positions[end_point],
|
||||
str(experiment.Conditions.WeatherId) + '_'
|
||||
+ str(experiment.id) + '_' + str(start_point)
|
||||
+ '.' + str(end_point))
|
||||
|
||||
# compute stats for the experiment
|
||||
|
||||
self._write_summary_results(
|
||||
experiment, pose, rep, euclidean_distance,
|
||||
remaining_distance, final_time, time_out, result)
|
||||
|
||||
self._write_details_results(experiment, rep, reward_vec)
|
||||
|
||||
if(result > 0):
|
||||
logging.info('+++++ Target achieved in %f seconds! +++++',
|
||||
final_time)
|
||||
else:
|
||||
logging.info('----- Timeout! -----')
|
||||
return self.get_all_statistics()
|
||||
|
||||
def _write_summary_results(self, experiment, pose, rep,
|
||||
path_distance, remaining_distance,
|
||||
final_time, time_out, result):
|
||||
|
||||
self._dict_stats['exp_id'] = experiment.id
|
||||
self._dict_stats['rep'] = rep
|
||||
self._dict_stats['weather'] = experiment.Conditions.WeatherId
|
||||
self._dict_stats['start_point'] = pose[0]
|
||||
self._dict_stats['end_point'] = pose[1]
|
||||
self._dict_stats['result'] = result
|
||||
self._dict_stats['initial_distance'] = path_distance
|
||||
self._dict_stats['final_distance'] = remaining_distance
|
||||
self._dict_stats['final_time'] = final_time
|
||||
self._dict_stats['time_out'] = time_out
|
||||
|
||||
with open(os.path.join(self._full_name, self._suffix_name), 'a+') as ofd:
|
||||
|
||||
w = csv.DictWriter(ofd, self._dict_stats.keys())
|
||||
|
||||
w.writerow(self._dict_stats)
|
||||
|
||||
def _write_details_results(self, experiment, rep, reward_vec):
|
||||
|
||||
with open(os.path.join(self._full_name,
|
||||
'details_' + self._suffix_name), 'a+') as rfd:
|
||||
|
||||
rw = csv.DictWriter(rfd, self._dict_rewards.keys())
|
||||
|
||||
for i in range(len(reward_vec)):
|
||||
self._dict_rewards['exp_id'] = experiment.id
|
||||
self._dict_rewards['rep'] = rep
|
||||
self._dict_rewards['weather'] = experiment.Conditions.WeatherId
|
||||
self._dict_rewards['collision_gen'] = reward_vec[
|
||||
i].collision_other
|
||||
self._dict_rewards['collision_ped'] = reward_vec[
|
||||
i].collision_pedestrians
|
||||
self._dict_rewards['collision_car'] = reward_vec[
|
||||
i].collision_vehicles
|
||||
self._dict_rewards['lane_intersect'] = reward_vec[
|
||||
i].intersection_otherlane
|
||||
self._dict_rewards['sidewalk_intersect'] = reward_vec[
|
||||
i].intersection_offroad
|
||||
self._dict_rewards['pos_x'] = reward_vec[
|
||||
i].transform.location.x
|
||||
self._dict_rewards['pos_y'] = reward_vec[
|
||||
i].transform.location.y
|
||||
|
||||
rw.writerow(self._dict_rewards)
|
||||
|
||||
def _create_log_record(self, base_name, experiments):
|
||||
"""
|
||||
This function creates the log files for the benchmark.
|
||||
|
||||
"""
|
||||
suffix_name = self._get_experiments_names(experiments)
|
||||
full_name = os.path.join('_benchmarks_results',
|
||||
base_name + '_'
|
||||
+ self._get_details() + '/')
|
||||
|
||||
folder = os.path.dirname(full_name)
|
||||
if not os.path.isdir(folder):
|
||||
os.makedirs(folder)
|
||||
|
||||
# Make a date file: to show when this was modified,
|
||||
# the number of times the experiments were run
|
||||
now = datetime.datetime.now()
|
||||
open(os.path.join(full_name, now.strftime("%Y%m%d%H%M")),'w').close()
|
||||
|
||||
return suffix_name, full_name
|
||||
|
||||
|
||||
def _continue_experiment(self, continue_experiment):
|
||||
|
||||
if self._experiment_exist():
|
||||
|
||||
if continue_experiment:
|
||||
line_on_file = self._get_last_position()
|
||||
|
||||
else:
|
||||
# Ask question, to avoid mistaken override situations
|
||||
answer = input_data("The experiment was already found in the files"
|
||||
+ ", Do you want to continue (y/n)? \n"
|
||||
)
|
||||
if answer == 'Yes' or answer == 'y':
|
||||
line_on_file = self._get_last_position()
|
||||
else:
|
||||
line_on_file = 0
|
||||
|
||||
else:
|
||||
line_on_file = 0
|
||||
|
||||
return line_on_file
|
||||
|
||||
|
||||
|
||||
def _experiment_exist(self):
|
||||
return os.path.isfile(self._full_name)
|
||||
|
||||
def _get_last_position(self):
|
||||
|
||||
with open(os.path.join(self._full_name, self._suffix_name)) as f:
|
||||
return sum(1 for _ in f)
|
||||
|
||||
|
||||
# To be redefined on subclasses on how to calculate timeout for an episode
|
||||
@abc.abstractmethod
|
||||
def _calculate_time_out(self, distance):
|
||||
pass
|
||||
|
||||
@abc.abstractmethod
|
||||
def _get_details(self):
|
||||
"""
|
||||
Get details
|
||||
:return: a string with name and town of the subclass
|
||||
"""
|
||||
@abc.abstractmethod
|
||||
def _build_experiments(self):
|
||||
"""
|
||||
Returns a set of experiments to be evaluated
|
||||
Must be redefined in an inherited class.
|
||||
|
||||
"""
|
||||
|
||||
@abc.abstractmethod
|
||||
def get_all_statistics(self):
|
||||
"""
|
||||
Get the statistics of the evaluated experiments
|
||||
:return:
|
||||
"""
|
||||
|
||||
@abc.abstractmethod
|
||||
def _get_pose_and_task(self, line_on_file):
|
||||
"""
|
||||
Parse the experiment depending on number of poses and tasks
|
||||
"""
|
||||
|
||||
|
||||
@abc.abstractmethod
|
||||
def plot_summary_train(self):
|
||||
"""
|
||||
returns the summary for the train weather/task episodes
|
||||
|
||||
"""
|
||||
|
||||
@abc.abstractmethod
|
||||
def plot_summary_test(self):
|
||||
"""
|
||||
returns the summary for the test weather/task episodes
|
||||
|
||||
"""
|
||||
@staticmethod
|
||||
def _get_experiments_names(experiments):
|
||||
|
||||
name_cat = 'w'
|
||||
|
||||
for experiment in experiments:
|
||||
|
||||
name_cat += str(experiment.Conditions.WeatherId) + '.'
|
||||
|
||||
return name_cat
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,203 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
# CORL experiment set.
|
||||
|
||||
from __future__ import print_function
|
||||
|
||||
import os
|
||||
|
||||
from .benchmark import Benchmark
|
||||
from .experiment import Experiment
|
||||
from carla.sensor import Camera
|
||||
from carla.settings import CarlaSettings
|
||||
|
||||
from .metrics import compute_summary
|
||||
|
||||
|
||||
class CoRL2017(Benchmark):
|
||||
|
||||
def get_all_statistics(self):
|
||||
|
||||
summary = compute_summary(os.path.join(
|
||||
self._full_name, self._suffix_name), [3])
|
||||
|
||||
return summary
|
||||
|
||||
def plot_summary_train(self):
|
||||
|
||||
self._plot_summary([1.0, 3.0, 6.0, 8.0])
|
||||
|
||||
def plot_summary_test(self):
|
||||
|
||||
self._plot_summary([4.0, 14.0])
|
||||
|
||||
def _plot_summary(self, weathers):
|
||||
"""
|
||||
We plot the summary of the testing for the set selected weathers.
|
||||
The test weathers are [4,14]
|
||||
|
||||
"""
|
||||
|
||||
metrics_summary = compute_summary(os.path.join(
|
||||
self._full_name, self._suffix_name), [3])
|
||||
|
||||
for metric, values in metrics_summary.items():
|
||||
|
||||
print('Metric : ', metric)
|
||||
for weather, tasks in values.items():
|
||||
if weather in set(weathers):
|
||||
print(' Weather: ', weather)
|
||||
count = 0
|
||||
for t in tasks:
|
||||
print(' Task ', count, ' -> ', t)
|
||||
count += 1
|
||||
|
||||
print(' AvG -> ', float(sum(tasks)) / float(len(tasks)))
|
||||
|
||||
def _calculate_time_out(self, distance):
|
||||
"""
|
||||
Function to return the timeout ( in miliseconds) that is calculated based on distance to goal.
|
||||
This is the same timeout as used on the CoRL paper.
|
||||
"""
|
||||
|
||||
return ((distance / 100000.0) / 10.0) * 3600.0 + 10.0
|
||||
|
||||
def _poses_town01(self):
|
||||
"""
|
||||
Each matrix is a new task. We have all the four tasks
|
||||
|
||||
"""
|
||||
|
||||
def _poses_straight():
|
||||
return [[36, 40], [39, 35], [110, 114], [7, 3], [0, 4],
|
||||
[68, 50], [61, 59], [47, 64], [147, 90], [33, 87],
|
||||
[26, 19], [80, 76], [45, 49], [55, 44], [29, 107],
|
||||
[95, 104], [84, 34], [53, 67], [22, 17], [91, 148],
|
||||
[20, 107], [78, 70], [95, 102], [68, 44], [45, 69]]
|
||||
|
||||
def _poses_one_curve():
|
||||
return [[138, 17], [47, 16], [26, 9], [42, 49], [140, 124],
|
||||
[85, 98], [65, 133], [137, 51], [76, 66], [46, 39],
|
||||
[40, 60], [0, 29], [4, 129], [121, 140], [2, 129],
|
||||
[78, 44], [68, 85], [41, 102], [95, 70], [68, 129],
|
||||
[84, 69], [47, 79], [110, 15], [130, 17], [0, 17]]
|
||||
|
||||
def _poses_navigation():
|
||||
return [[105, 29], [27, 130], [102, 87], [132, 27], [24, 44],
|
||||
[96, 26], [34, 67], [28, 1], [140, 134], [105, 9],
|
||||
[148, 129], [65, 18], [21, 16], [147, 97], [42, 51],
|
||||
[30, 41], [18, 107], [69, 45], [102, 95], [18, 145],
|
||||
[111, 64], [79, 45], [84, 69], [73, 31], [37, 81]]
|
||||
|
||||
return [_poses_straight(),
|
||||
_poses_one_curve(),
|
||||
_poses_navigation(),
|
||||
_poses_navigation()]
|
||||
|
||||
def _poses_town02(self):
|
||||
|
||||
def _poses_straight():
|
||||
return [[38, 34], [4, 2], [12, 10], [62, 55], [43, 47],
|
||||
[64, 66], [78, 76], [59, 57], [61, 18], [35, 39],
|
||||
[12, 8], [0, 18], [75, 68], [54, 60], [45, 49],
|
||||
[46, 42], [53, 46], [80, 29], [65, 63], [0, 81],
|
||||
[54, 63], [51, 42], [16, 19], [17, 26], [77, 68]]
|
||||
|
||||
def _poses_one_curve():
|
||||
return [[37, 76], [8, 24], [60, 69], [38, 10], [21, 1],
|
||||
[58, 71], [74, 32], [44, 0], [71, 16], [14, 24],
|
||||
[34, 11], [43, 14], [75, 16], [80, 21], [3, 23],
|
||||
[75, 59], [50, 47], [11, 19], [77, 34], [79, 25],
|
||||
[40, 63], [58, 76], [79, 55], [16, 61], [27, 11]]
|
||||
|
||||
def _poses_navigation():
|
||||
return [[19, 66], [79, 14], [19, 57], [23, 1],
|
||||
[53, 76], [42, 13], [31, 71], [33, 5],
|
||||
[54, 30], [10, 61], [66, 3], [27, 12],
|
||||
[79, 19], [2, 29], [16, 14], [5, 57],
|
||||
[70, 73], [46, 67], [57, 50], [61, 49], [21, 12],
|
||||
[51, 81], [77, 68], [56, 65], [43, 54]]
|
||||
|
||||
return [_poses_straight(),
|
||||
_poses_one_curve(),
|
||||
_poses_navigation(),
|
||||
_poses_navigation()
|
||||
]
|
||||
|
||||
def _build_experiments(self):
|
||||
"""
|
||||
Creates the whole set of experiment objects,
|
||||
The experiments created depend on the selected Town.
|
||||
"""
|
||||
|
||||
# We set the camera
|
||||
# This single RGB camera is used on every experiment
|
||||
|
||||
camera = Camera('CameraRGB')
|
||||
camera.set(CameraFOV=100)
|
||||
|
||||
camera.set_image_size(800, 600)
|
||||
|
||||
camera.set_position(200, 0, 140)
|
||||
camera.set_rotation(-15.0, 0, 0)
|
||||
|
||||
weathers = [1, 3, 6, 8, 4, 14]
|
||||
if self._city_name == 'Town01':
|
||||
poses_tasks = self._poses_town01()
|
||||
vehicles_tasks = [0, 0, 0, 20]
|
||||
pedestrians_tasks = [0, 0, 0, 50]
|
||||
else:
|
||||
poses_tasks = self._poses_town02()
|
||||
vehicles_tasks = [0, 0, 0, 15]
|
||||
pedestrians_tasks = [0, 0, 0, 50]
|
||||
|
||||
experiments_vector = []
|
||||
|
||||
for weather in weathers:
|
||||
|
||||
for iteration in range(len(poses_tasks)):
|
||||
poses = poses_tasks[iteration]
|
||||
vehicles = vehicles_tasks[iteration]
|
||||
pedestrians = pedestrians_tasks[iteration]
|
||||
|
||||
conditions = CarlaSettings()
|
||||
conditions.set(
|
||||
SynchronousMode=True,
|
||||
SendNonPlayerAgentsInfo=True,
|
||||
NumberOfVehicles=vehicles,
|
||||
NumberOfPedestrians=pedestrians,
|
||||
WeatherId=weather,
|
||||
SeedVehicles=123456789,
|
||||
SeedPedestrians=123456789
|
||||
)
|
||||
# Add all the cameras that were set for this experiments
|
||||
|
||||
conditions.add_sensor(camera)
|
||||
|
||||
experiment = Experiment()
|
||||
experiment.set(
|
||||
Conditions=conditions,
|
||||
Poses=poses,
|
||||
Id=iteration,
|
||||
Repetitions=1
|
||||
)
|
||||
experiments_vector.append(experiment)
|
||||
|
||||
return experiments_vector
|
||||
|
||||
def _get_details(self):
|
||||
|
||||
# Function to get automatic information from the experiment for writing purposes
|
||||
return 'corl2017_' + self._city_name
|
||||
|
||||
def _get_pose_and_task(self, line_on_file):
|
||||
"""
|
||||
Returns the pose and task this experiment is, based on the line it was
|
||||
on the log file.
|
||||
"""
|
||||
# We assume that the number of poses is constant
|
||||
return int(line_on_file / len(self._experiments)), line_on_file % 25
|
||||
|
|
@ -0,0 +1,38 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
from carla.settings import CarlaSettings
|
||||
|
||||
|
||||
class Experiment(object):
|
||||
|
||||
def __init__(self):
|
||||
self.Id = ''
|
||||
self.Conditions = CarlaSettings()
|
||||
self.Poses = [[]]
|
||||
self.Repetitions = 1
|
||||
|
||||
def set(self, **kwargs):
|
||||
for key, value in kwargs.items():
|
||||
if not hasattr(self, key):
|
||||
raise ValueError('Experiment: no key named %r' % key)
|
||||
setattr(self, key, value)
|
||||
|
||||
@property
|
||||
def id(self):
|
||||
return self.Id
|
||||
|
||||
@property
|
||||
def conditions(self):
|
||||
return self.Conditions
|
||||
|
||||
@property
|
||||
def poses(self):
|
||||
return self.Poses
|
||||
|
||||
@property
|
||||
def repetitions(self):
|
||||
return self.Repetitions
|
||||
|
|
@ -0,0 +1,205 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
|
||||
import numpy as np
|
||||
import math
|
||||
import os
|
||||
|
||||
|
||||
sldist = lambda c1, c2: math.sqrt((c2[0] - c1[0])**2 + (c2[1] - c1[1])**2)
|
||||
flatten = lambda l: [item for sublist in l for item in sublist]
|
||||
|
||||
|
||||
def get_colisions(selected_matrix, header):
|
||||
|
||||
count_gen = 0
|
||||
count_ped = 0
|
||||
count_car = 0
|
||||
i = 1
|
||||
|
||||
while i < selected_matrix.shape[0]:
|
||||
if (selected_matrix[i, header.index('collision_gen')]
|
||||
- selected_matrix[(i-10), header.index('collision_gen')]) > 40000:
|
||||
count_gen += 1
|
||||
i += 20
|
||||
i += 1
|
||||
|
||||
i = 1
|
||||
while i < selected_matrix.shape[0]:
|
||||
if (selected_matrix[i, header.index('collision_car')]
|
||||
- selected_matrix[(i-10), header.index('collision_car')]) > 40000:
|
||||
count_car += 1
|
||||
i += 30
|
||||
i += 1
|
||||
|
||||
i = 1
|
||||
while i < selected_matrix.shape[0]:
|
||||
if (selected_matrix[i, header.index('collision_ped')]
|
||||
- selected_matrix[i-5, header.index('collision_ped')]) > 30000:
|
||||
count_ped += 1
|
||||
i += 100
|
||||
i += 1
|
||||
|
||||
return count_gen, count_car, count_ped
|
||||
|
||||
|
||||
def get_distance_traveled(selected_matrix, header):
|
||||
|
||||
prev_x = selected_matrix[0, header.index('pos_x')]
|
||||
prev_y = selected_matrix[0, header.index('pos_y')]
|
||||
|
||||
i = 1
|
||||
acummulated_distance = 0
|
||||
while i < selected_matrix.shape[0]:
|
||||
|
||||
x = selected_matrix[i, header.index('pos_x')]
|
||||
y = selected_matrix[i, header.index('pos_y')]
|
||||
# Here we defined a maximun distance in a tick, this case 8 meters or 288km/h
|
||||
if sldist((x, y), (prev_x, prev_y)) < 800:
|
||||
acummulated_distance += sldist((x, y), (prev_x, prev_y))
|
||||
|
||||
|
||||
prev_x = x
|
||||
prev_y = y
|
||||
|
||||
i += 1
|
||||
|
||||
return float(acummulated_distance)/float(100*1000)
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
def get_out_of_road_lane(selected_matrix, header):
|
||||
|
||||
count_road = 0
|
||||
count_lane = 0
|
||||
|
||||
i = 0
|
||||
|
||||
while i < selected_matrix.shape[0]:
|
||||
# print selected_matrix[i,6]
|
||||
if (selected_matrix[i, header.index('sidewalk_intersect')]
|
||||
- selected_matrix[(i-10), header.index('sidewalk_intersect')]) > 0.3:
|
||||
count_road += 1
|
||||
i += 20
|
||||
if i >= selected_matrix.shape[0]:
|
||||
break
|
||||
|
||||
if (selected_matrix[i, header.index('lane_intersect')]
|
||||
- selected_matrix[(i-10), header.index('lane_intersect')]) > 0.4:
|
||||
count_lane += 1
|
||||
i += 20
|
||||
|
||||
i += 1
|
||||
|
||||
return count_lane, count_road
|
||||
|
||||
|
||||
|
||||
def compute_summary(filename, dynamic_episodes):
|
||||
|
||||
# Separate the PATH and the basename
|
||||
path = os.path.dirname(filename)
|
||||
base_name = os.path.basename(filename)
|
||||
|
||||
|
||||
|
||||
f = open(filename, "rb")
|
||||
header = f.readline()
|
||||
header = header.split(',')
|
||||
header[-1] = header[-1][:-2]
|
||||
f.close()
|
||||
|
||||
f = open(os.path.join(path, 'details_' + base_name), "rb")
|
||||
header_details = f.readline()
|
||||
header_details = header_details.split(',')
|
||||
header_details[-1] = header_details[-1][:-2]
|
||||
f.close()
|
||||
|
||||
data_matrix = np.loadtxt(open(filename, "rb"), delimiter=",", skiprows=1)
|
||||
|
||||
# Corner Case: The presented test just had one episode
|
||||
if data_matrix.ndim == 1:
|
||||
data_matrix = np.expand_dims(data_matrix, axis=0)
|
||||
|
||||
|
||||
tasks = np.unique(data_matrix[:, header.index('exp_id')])
|
||||
|
||||
all_weathers = np.unique(data_matrix[:, header.index('weather')])
|
||||
|
||||
reward_matrix = np.loadtxt(open(os.path.join(
|
||||
path, 'details_' + base_name), "rb"), delimiter=",", skiprows=1)
|
||||
|
||||
metrics_dictionary = {'average_completion': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'intersection_offroad': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'intersection_otherlane': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'collision_pedestrians': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'collision_vehicles': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'collision_other': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'average_fully_completed': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'average_speed': {w: [0.0]*len(tasks) for w in all_weathers},
|
||||
'driven_kilometers': {w: [0.0]*len(tasks) for w in all_weathers}
|
||||
}
|
||||
|
||||
for t in tasks:
|
||||
task_data_matrix = data_matrix[
|
||||
data_matrix[:, header.index('exp_id')] == t]
|
||||
weathers = np.unique(task_data_matrix[:, header.index('weather')])
|
||||
|
||||
|
||||
|
||||
for w in weathers:
|
||||
t = int(t)
|
||||
|
||||
task_data_matrix = data_matrix[np.logical_and(data_matrix[:, header.index(
|
||||
'exp_id')] == t, data_matrix[:, header.index('weather')] == w)]
|
||||
|
||||
|
||||
task_reward_matrix = reward_matrix[np.logical_and(reward_matrix[:, header_details.index(
|
||||
'exp_id')] == float(t), reward_matrix[:, header_details.index('weather')] == float(w))]
|
||||
|
||||
km_run = get_distance_traveled(
|
||||
task_reward_matrix, header_details)
|
||||
|
||||
metrics_dictionary['average_fully_completed'][w][t] = sum(
|
||||
task_data_matrix[:, header.index('result')])/task_data_matrix.shape[0]
|
||||
|
||||
metrics_dictionary['average_completion'][w][t] = sum(
|
||||
(task_data_matrix[:, header.index('initial_distance')]
|
||||
- task_data_matrix[:, header.index('final_distance')])
|
||||
/ task_data_matrix[:, header.index('initial_distance')]) \
|
||||
/ len(task_data_matrix[:, header.index('final_distance')])
|
||||
|
||||
|
||||
metrics_dictionary['driven_kilometers'][w][t]= km_run
|
||||
metrics_dictionary['average_speed'][w][t]= km_run/ \
|
||||
((sum(task_data_matrix[:, header.index('final_time')]))/3600.0)
|
||||
|
||||
|
||||
|
||||
if list(tasks).index(t) in set(dynamic_episodes):
|
||||
|
||||
lane_road = get_out_of_road_lane(
|
||||
task_reward_matrix, header_details)
|
||||
colisions = get_colisions(task_reward_matrix, header_details)
|
||||
|
||||
|
||||
|
||||
metrics_dictionary['intersection_offroad'][
|
||||
w][t] = lane_road[0]/km_run
|
||||
metrics_dictionary['intersection_otherlane'][
|
||||
w][t] = lane_road[1]/km_run
|
||||
metrics_dictionary['collision_pedestrians'][
|
||||
w][t] = colisions[2]/km_run
|
||||
metrics_dictionary['collision_vehicles'][
|
||||
w][t] = colisions[1]/km_run
|
||||
metrics_dictionary['collision_other'][
|
||||
w][t] = colisions[0]/km_run
|
||||
|
||||
|
||||
return metrics_dictionary
|
||||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -171,20 +171,28 @@ class CarlaClient(object):
|
|||
|
||||
@staticmethod
|
||||
def _iterate_sensor_data(raw_data):
|
||||
# At this point the only sensors available are images, the raw_data
|
||||
# consists of images only.
|
||||
image_types = ['None', 'SceneFinal', 'Depth', 'SemanticSegmentation']
|
||||
image_types = ['None', 'SceneFinal', 'Depth', 'SemanticSegmentation', 'Lidar']
|
||||
gettype = lambda id: image_types[id] if len(image_types) > id else 'Unknown'
|
||||
getint = lambda index: struct.unpack('<L', raw_data[index*4:index*4+4])[0]
|
||||
getfloat = lambda index: struct.unpack('<f', raw_data[index*4:index*4+4])[0]
|
||||
getdouble = lambda index: struct.unpack('<d', raw_data[index*4:index*4+8])[0]
|
||||
total_size = len(raw_data) / 4
|
||||
index = 0
|
||||
while index < total_size:
|
||||
width = getint(index)
|
||||
height = getint(index + 1)
|
||||
image_type = gettype(getint(index + 2))
|
||||
fov = getfloat(index + 3)
|
||||
begin = index + 4
|
||||
end = begin + width * height
|
||||
index = end
|
||||
yield sensor.Image(width, height, image_type, fov, raw_data[begin*4:end*4])
|
||||
sensor_type = gettype(getval(index + 2))
|
||||
if sensor_type == 'Lidar':
|
||||
horizontal_angle = getdouble(index)
|
||||
channels_count = getint(index + 3)
|
||||
lm = sensor.LidarMeasurement(
|
||||
horizontal_angle, channels_count,
|
||||
sensor_type, raw_data[index*4:])
|
||||
index += lm.size_in_bytes
|
||||
yield lm
|
||||
else:
|
||||
width = getint(index)
|
||||
height = getint(index + 1)
|
||||
fov = getfloat(index + 3)
|
||||
begin = index + 4
|
||||
end = begin + width * height
|
||||
index = end
|
||||
yield sensor.Image(width, height, sensor_type, fov, raw_data[begin*4:end*4])
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
Binary file not shown.
|
After Width: | Height: | Size: 27 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 10 KiB |
|
|
@ -0,0 +1,156 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import heapq
|
||||
|
||||
|
||||
class Cell(object):
|
||||
def __init__(self, x, y, reachable):
|
||||
"""Initialize new cell.
|
||||
|
||||
@param reachable is cell reachable? not a wall?
|
||||
@param x cell x coordinate
|
||||
@param y cell y coordinate
|
||||
@param g cost to move from the starting cell to this cell.
|
||||
@param h estimation of the cost to move from this cell
|
||||
to the ending cell.
|
||||
@param f f = g + h
|
||||
"""
|
||||
self.reachable = reachable
|
||||
self.x = x
|
||||
self.y = y
|
||||
self.parent = None
|
||||
self.g = 0
|
||||
self.h = 0
|
||||
self.f = 0
|
||||
|
||||
def __lt__(self, other):
|
||||
return self.g < other.g
|
||||
|
||||
|
||||
class AStar(object):
|
||||
def __init__(self):
|
||||
# open list
|
||||
self.opened = []
|
||||
heapq.heapify(self.opened)
|
||||
# visited cells list
|
||||
self.closed = set()
|
||||
# grid cells
|
||||
self.cells = []
|
||||
self.grid_height = None
|
||||
self.grid_width = None
|
||||
self.start = None
|
||||
self.end = None
|
||||
|
||||
def init_grid(self, width, height, walls, start, end):
|
||||
"""Prepare grid cells, walls.
|
||||
|
||||
@param width grid's width.
|
||||
@param height grid's height.
|
||||
@param walls list of wall x,y tuples.
|
||||
@param start grid starting point x,y tuple.
|
||||
@param end grid ending point x,y tuple.
|
||||
"""
|
||||
self.grid_height = height
|
||||
self.grid_width = width
|
||||
for x in range(self.grid_width):
|
||||
for y in range(self.grid_height):
|
||||
if (x, y) in walls:
|
||||
reachable = False
|
||||
else:
|
||||
reachable = True
|
||||
self.cells.append(Cell(x, y, reachable))
|
||||
self.start = self.get_cell(*start)
|
||||
self.end = self.get_cell(*end)
|
||||
|
||||
def get_heuristic(self, cell):
|
||||
"""Compute the heuristic value H for a cell.
|
||||
|
||||
Distance between this cell and the ending cell multiply by 10.
|
||||
|
||||
@returns heuristic value H
|
||||
"""
|
||||
return 10 * (abs(cell.x - self.end.x) + abs(cell.y - self.end.y))
|
||||
|
||||
def get_cell(self, x, y):
|
||||
"""Returns a cell from the cells list.
|
||||
|
||||
@param x cell x coordinate
|
||||
@param y cell y coordinate
|
||||
@returns cell
|
||||
"""
|
||||
return self.cells[x * self.grid_height + y]
|
||||
|
||||
def get_adjacent_cells(self, cell):
|
||||
"""Returns adjacent cells to a cell.
|
||||
|
||||
Clockwise starting from the one on the right.
|
||||
|
||||
@param cell get adjacent cells for this cell
|
||||
@returns adjacent cells list.
|
||||
"""
|
||||
cells = []
|
||||
if cell.x < self.grid_width - 1:
|
||||
cells.append(self.get_cell(cell.x + 1, cell.y))
|
||||
if cell.y > 0:
|
||||
cells.append(self.get_cell(cell.x, cell.y - 1))
|
||||
if cell.x > 0:
|
||||
cells.append(self.get_cell(cell.x - 1, cell.y))
|
||||
if cell.y < self.grid_height - 1:
|
||||
cells.append(self.get_cell(cell.x, cell.y + 1))
|
||||
return cells
|
||||
|
||||
def get_path(self):
|
||||
cell = self.end
|
||||
path = [(cell.x, cell.y)]
|
||||
while cell.parent is not self.start:
|
||||
cell = cell.parent
|
||||
path.append((cell.x, cell.y))
|
||||
|
||||
path.append((self.start.x, self.start.y))
|
||||
path.reverse()
|
||||
return path
|
||||
|
||||
def update_cell(self, adj, cell):
|
||||
"""Update adjacent cell.
|
||||
|
||||
@param adj adjacent cell to current cell
|
||||
@param cell current cell being processed
|
||||
"""
|
||||
adj.g = cell.g + 10
|
||||
adj.h = self.get_heuristic(adj)
|
||||
adj.parent = cell
|
||||
adj.f = adj.h + adj.g
|
||||
|
||||
def solve(self):
|
||||
"""Solve maze, find path to ending cell.
|
||||
|
||||
@returns path or None if not found.
|
||||
"""
|
||||
# add starting cell to open heap queue
|
||||
heapq.heappush(self.opened, (self.start.f, self.start))
|
||||
while len(self.opened):
|
||||
# pop cell from heap queue
|
||||
_, cell = heapq.heappop(self.opened)
|
||||
# add cell to closed list so we don't process it twice
|
||||
self.closed.add(cell)
|
||||
# if ending cell, return found path
|
||||
if cell is self.end:
|
||||
return self.get_path()
|
||||
# get adjacent cells for cell
|
||||
adj_cells = self.get_adjacent_cells(cell)
|
||||
for adj_cell in adj_cells:
|
||||
if adj_cell.reachable and adj_cell not in self.closed:
|
||||
if (adj_cell.f, adj_cell) in self.opened:
|
||||
# if adj cell in open list, check if current path is
|
||||
# better than the one previously found
|
||||
# for this adj cell.
|
||||
if adj_cell.g > cell.g + 10:
|
||||
self.update_cell(adj_cell, cell)
|
||||
else:
|
||||
self.update_cell(adj_cell, cell)
|
||||
# add adj cell to open list
|
||||
heapq.heappush(self.opened, (adj_cell.f, adj_cell))
|
||||
|
|
@ -0,0 +1,136 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
from carla.planner.graph import sldist
|
||||
|
||||
from carla.planner.astar import AStar
|
||||
from carla.planner.map import CarlaMap
|
||||
|
||||
|
||||
class CityTrack(object):
|
||||
|
||||
def __init__(self, city_name):
|
||||
|
||||
self._node_density = 50.0
|
||||
self._pixel_density = 16.43
|
||||
|
||||
self._map = CarlaMap(city_name, self._pixel_density, self._node_density)
|
||||
|
||||
self._astar = AStar()
|
||||
|
||||
# Refers to the start position of the previous route computation
|
||||
self._previous_node = []
|
||||
|
||||
# The current computed route
|
||||
self._route = None
|
||||
|
||||
def project_node(self, position):
|
||||
"""
|
||||
Projecting the graph node into the city road
|
||||
"""
|
||||
|
||||
node = self._map.convert_to_node(position)
|
||||
|
||||
# To change the orientation with respect to the map standards
|
||||
|
||||
node = tuple([int(x) for x in node])
|
||||
|
||||
# Set to zero if it is less than zero.
|
||||
|
||||
node = (max(0, node[0]), max(0, node[1]))
|
||||
node = (min(self._map.get_graph_resolution()[0] - 1, node[0]),
|
||||
min(self._map.get_graph_resolution()[1] - 1, node[1]))
|
||||
|
||||
node = self._map.search_on_grid(node)
|
||||
|
||||
return node
|
||||
|
||||
def get_intersection_nodes(self):
|
||||
return self._map.get_intersection_nodes()
|
||||
|
||||
def get_pixel_density(self):
|
||||
return self._pixel_density
|
||||
|
||||
def get_node_density(self):
|
||||
return self._node_density
|
||||
|
||||
def is_at_goal(self, source, target):
|
||||
return source == target
|
||||
|
||||
def is_at_new_node(self, current_node):
|
||||
return current_node != self._previous_node
|
||||
|
||||
def is_away_from_intersection(self, current_node):
|
||||
return self._closest_intersection_position(current_node) > 1
|
||||
|
||||
def is_far_away_from_route_intersection(self, current_node):
|
||||
# CHECK FOR THE EMPTY CASE
|
||||
if self._route is None:
|
||||
raise RuntimeError('Impossible to find route'
|
||||
+ ' Current planner is limited'
|
||||
+ ' Try to select start points away from intersections')
|
||||
|
||||
return self._closest_intersection_route_position(current_node,
|
||||
self._route) > 4
|
||||
|
||||
def compute_route(self, node_source, source_ori, node_target, target_ori):
|
||||
|
||||
self._previous_node = node_source
|
||||
|
||||
a_star = AStar()
|
||||
a_star.init_grid(self._map.get_graph_resolution()[0],
|
||||
self._map.get_graph_resolution()[1],
|
||||
self._map.get_walls_directed(node_source, source_ori,
|
||||
node_target, target_ori), node_source,
|
||||
node_target)
|
||||
|
||||
route = a_star.solve()
|
||||
|
||||
# JuSt a Corner Case
|
||||
# Clean this to avoid having to use this function
|
||||
if route is None:
|
||||
a_star = AStar()
|
||||
a_star.init_grid(self._map.get_graph_resolution()[0],
|
||||
self._map.get_graph_resolution()[1], self._map.get_walls(),
|
||||
node_source, node_target)
|
||||
|
||||
route = a_star.solve()
|
||||
|
||||
self._route = route
|
||||
|
||||
return route
|
||||
|
||||
def get_distance_closest_node_route(self, pos, route):
|
||||
distance = []
|
||||
|
||||
for node_iter in route:
|
||||
|
||||
if node_iter in self._map.get_intersection_nodes():
|
||||
distance.append(sldist(node_iter, pos))
|
||||
|
||||
if not distance:
|
||||
return sldist(route[-1], pos)
|
||||
return sorted(distance)[0]
|
||||
|
||||
|
||||
def _closest_intersection_position(self, current_node):
|
||||
|
||||
distance_vector = []
|
||||
for node_iterator in self._map.get_intersection_nodes():
|
||||
distance_vector.append(sldist(node_iterator, current_node))
|
||||
|
||||
return sorted(distance_vector)[0]
|
||||
|
||||
|
||||
def _closest_intersection_route_position(self, current_node, route):
|
||||
|
||||
distance_vector = []
|
||||
for _ in route:
|
||||
for node_iterator in self._map.get_intersection_nodes():
|
||||
distance_vector.append(sldist(node_iterator, current_node))
|
||||
|
||||
return sorted(distance_vector)[0]
|
||||
|
||||
|
|
@ -0,0 +1,166 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import math
|
||||
import numpy as np
|
||||
|
||||
from carla.planner.graph import string_to_floats
|
||||
|
||||
# Constant definition enumeration
|
||||
|
||||
PIXEL = 0
|
||||
WORLD = 1
|
||||
NODE = 2
|
||||
|
||||
|
||||
class Converter(object):
|
||||
|
||||
def __init__(self, city_file, pixel_density, node_density):
|
||||
|
||||
self._node_density = node_density
|
||||
self._pixel_density = pixel_density
|
||||
with open(city_file, 'r') as f:
|
||||
# The offset of the world from the zero coordinates ( The
|
||||
# coordinate we consider zero)
|
||||
self._worldoffset = string_to_floats(f.readline())
|
||||
|
||||
angles = string_to_floats(f.readline())
|
||||
|
||||
# If there is an rotation between the world and map coordinates.
|
||||
self._worldrotation = np.array([
|
||||
[math.cos(math.radians(angles[2])), -math.sin(math.radians(angles[2])), 0.0],
|
||||
[math.sin(math.radians(angles[2])), math.cos(math.radians(angles[2])), 0.0],
|
||||
[0.0, 0.0, 1.0]])
|
||||
|
||||
# Ignore for now, these are offsets for map coordinates and scale
|
||||
# (not used).
|
||||
_ = f.readline()
|
||||
|
||||
# The offset of the map zero coordinate.
|
||||
self._mapoffset = string_to_floats(f.readline())
|
||||
|
||||
def convert_to_node(self, input_data):
|
||||
"""
|
||||
Receives a data type (Can Be Pixel or World )
|
||||
:param input_data: position in some coordinate
|
||||
:return: A vector representing a node
|
||||
"""
|
||||
|
||||
input_type = self._check_input_type(input_data)
|
||||
if input_type == PIXEL:
|
||||
return self._pixel_to_node(input_data)
|
||||
elif input_type == WORLD:
|
||||
return self._world_to_node(input_data)
|
||||
else:
|
||||
raise ValueError('Invalid node to be converted')
|
||||
|
||||
def convert_to_pixel(self, input_data):
|
||||
|
||||
"""
|
||||
Receives a data type (Can Be Node or World )
|
||||
:param input_data: position in some coordinate
|
||||
:return: A vector with pixel coordinates
|
||||
"""
|
||||
|
||||
input_type = self._check_input_type(input_data)
|
||||
|
||||
if input_type == NODE:
|
||||
return self._node_to_pixel(input_data)
|
||||
elif input_type == WORLD:
|
||||
return self._world_to_pixel(input_data)
|
||||
else:
|
||||
raise ValueError('Invalid node to be converted')
|
||||
|
||||
def convert_to_world(self, input_data):
|
||||
|
||||
"""
|
||||
Receives a data type (Can Be Pixel or Node )
|
||||
:param input_data: position in some coordinate
|
||||
:return: vector with world coordinates
|
||||
"""
|
||||
|
||||
input_type = self._check_input_type(input_data)
|
||||
if input_type == NODE:
|
||||
return self._node_to_world(input_data)
|
||||
elif input_type == PIXEL:
|
||||
return self._pixel_to_world(input_data)
|
||||
else:
|
||||
raise ValueError('Invalid node to be converted')
|
||||
|
||||
def _node_to_pixel(self, node):
|
||||
"""
|
||||
Conversion from node format (graph) to pixel (image)
|
||||
:param node:
|
||||
:return: pixel
|
||||
"""
|
||||
pixel = [((node[0] + 2) * self._node_density)
|
||||
, ((node[1] + 2) * self._node_density)]
|
||||
return pixel
|
||||
|
||||
def _pixel_to_node(self, pixel):
|
||||
"""
|
||||
Conversion from pixel format (image) to node (graph)
|
||||
:param node:
|
||||
:return: pixel
|
||||
"""
|
||||
node = [int(((pixel[0]) / self._node_density) - 2)
|
||||
, int(((pixel[1]) / self._node_density) - 2)]
|
||||
|
||||
return tuple(node)
|
||||
|
||||
def _pixel_to_world(self, pixel):
|
||||
"""
|
||||
Conversion from pixel format (image) to world (3D)
|
||||
:param pixel:
|
||||
:return: world
|
||||
"""
|
||||
|
||||
relative_location = [pixel[0] * self._pixel_density,
|
||||
pixel[1] * self._pixel_density]
|
||||
|
||||
world = [
|
||||
relative_location[0] + self._mapoffset[0] - self._worldoffset[0],
|
||||
relative_location[1] + self._mapoffset[1] - self._worldoffset[1],
|
||||
22
|
||||
]
|
||||
|
||||
return world
|
||||
|
||||
def _world_to_pixel(self, world):
|
||||
"""
|
||||
Conversion from world format (3D) to pixel
|
||||
:param world:
|
||||
:return: pixel
|
||||
"""
|
||||
|
||||
rotation = np.array([world[0], world[1], world[2]])
|
||||
rotation = rotation.dot(self._worldrotation)
|
||||
|
||||
relative_location = [rotation[0] + self._worldoffset[0] - self._mapoffset[0],
|
||||
rotation[1] + self._worldoffset[1] - self._mapoffset[1],
|
||||
rotation[2] + self._worldoffset[2] - self._mapoffset[2]]
|
||||
|
||||
|
||||
|
||||
pixel = [math.floor(relative_location[0] / float(self._pixel_density)),
|
||||
math.floor(relative_location[1] / float(self._pixel_density))]
|
||||
|
||||
return pixel
|
||||
|
||||
def _world_to_node(self, world):
|
||||
return self._pixel_to_node(self._world_to_pixel(world))
|
||||
|
||||
def _node_to_world(self, node):
|
||||
|
||||
return self._pixel_to_world(self._node_to_pixel(node))
|
||||
|
||||
def _check_input_type(self, input_data):
|
||||
if len(input_data) > 2:
|
||||
return WORLD
|
||||
elif type(input_data[0]) is int:
|
||||
return NODE
|
||||
else:
|
||||
return PIXEL
|
||||
|
|
@ -0,0 +1,141 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import math
|
||||
import numpy as np
|
||||
|
||||
|
||||
def string_to_node(string):
|
||||
vec = string.split(',')
|
||||
return (int(vec[0]), int(vec[1]))
|
||||
|
||||
|
||||
def string_to_floats(string):
|
||||
vec = string.split(',')
|
||||
return (float(vec[0]), float(vec[1]), float(vec[2]))
|
||||
|
||||
|
||||
def sldist(c1, c2):
|
||||
return math.sqrt((c2[0] - c1[0]) ** 2 + (c2[1] - c1[1]) ** 2)
|
||||
|
||||
|
||||
def sldist3(c1, c2):
|
||||
return math.sqrt((c2[0] - c1[0]) ** 2 + (c2[1] - c1[1])
|
||||
** 2 + (c2[2] - c1[2]) ** 2)
|
||||
|
||||
|
||||
class Graph(object):
|
||||
"""
|
||||
A simple directed, weighted graph
|
||||
"""
|
||||
|
||||
def __init__(self, graph_file=None, node_density=50):
|
||||
|
||||
self._nodes = set()
|
||||
self._angles = {}
|
||||
self._edges = {}
|
||||
self._distances = {}
|
||||
self._node_density = node_density
|
||||
|
||||
if graph_file is not None:
|
||||
with open(graph_file, 'r') as f:
|
||||
# Skipe the first four lines that
|
||||
lines_after_4 = f.readlines()[4:]
|
||||
|
||||
# the graph resolution.
|
||||
linegraphres = lines_after_4[0]
|
||||
self._resolution = string_to_node(linegraphres)
|
||||
for line in lines_after_4[1:]:
|
||||
|
||||
from_node, to_node, d = line.split()
|
||||
from_node = string_to_node(from_node)
|
||||
to_node = string_to_node(to_node)
|
||||
|
||||
if from_node not in self._nodes:
|
||||
self.add_node(from_node)
|
||||
if to_node not in self._nodes:
|
||||
self.add_node(to_node)
|
||||
|
||||
self._edges.setdefault(from_node, [])
|
||||
self._edges[from_node].append(to_node)
|
||||
self._distances[(from_node, to_node)] = float(d)
|
||||
|
||||
def add_node(self, value):
|
||||
self._nodes.add(value)
|
||||
|
||||
def make_orientations(self, node, heading):
|
||||
|
||||
import collections
|
||||
distance_dic = {}
|
||||
for node_iter in self._nodes:
|
||||
if node_iter != node:
|
||||
distance_dic[sldist(node, node_iter)] = node_iter
|
||||
|
||||
distance_dic = collections.OrderedDict(
|
||||
sorted(distance_dic.items()))
|
||||
|
||||
self._angles[node] = heading
|
||||
for _, v in distance_dic.items():
|
||||
start_to_goal = np.array([node[0] - v[0], node[1] - v[1]])
|
||||
|
||||
print(start_to_goal)
|
||||
|
||||
self._angles[v] = start_to_goal / np.linalg.norm(start_to_goal)
|
||||
|
||||
def add_edge(self, from_node, to_node, distance):
|
||||
self._add_edge(from_node, to_node, distance)
|
||||
|
||||
def _add_edge(self, from_node, to_node, distance):
|
||||
self._edges.setdefault(from_node, [])
|
||||
self._edges[from_node].append(to_node)
|
||||
self._distances[(from_node, to_node)] = distance
|
||||
|
||||
def get_resolution(self):
|
||||
return self._resolution
|
||||
def get_edges(self):
|
||||
return self._edges
|
||||
|
||||
def intersection_nodes(self):
|
||||
|
||||
intersect_nodes = []
|
||||
for node in self._nodes:
|
||||
if len(self._edges[node]) > 2:
|
||||
intersect_nodes.append(node)
|
||||
|
||||
return intersect_nodes
|
||||
|
||||
# This contains also the non-intersection turns...
|
||||
|
||||
def turn_nodes(self):
|
||||
|
||||
return self._nodes
|
||||
|
||||
def plot_ori(self, c):
|
||||
from matplotlib import collections as mc
|
||||
|
||||
import matplotlib.pyplot as plt
|
||||
line_len = 1
|
||||
|
||||
lines = [[(p[0], p[1]), (p[0] + line_len * self._angles[p][0],
|
||||
p[1] + line_len * self._angles[p][1])] for p in self._nodes]
|
||||
lc = mc.LineCollection(lines, linewidth=2, color='green')
|
||||
_, ax = plt.subplots()
|
||||
ax.add_collection(lc)
|
||||
|
||||
ax.autoscale()
|
||||
ax.margins(0.1)
|
||||
|
||||
xs = [p[0] for p in self._nodes]
|
||||
ys = [p[1] for p in self._nodes]
|
||||
|
||||
plt.scatter(xs, ys, color=c)
|
||||
|
||||
def plot(self, c):
|
||||
import matplotlib.pyplot as plt
|
||||
xs = [p[0] for p in self._nodes]
|
||||
ys = [p[1] for p in self._nodes]
|
||||
|
||||
plt.scatter(xs, ys, color=c)
|
||||
|
|
@ -0,0 +1,135 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import copy
|
||||
import numpy as np
|
||||
|
||||
|
||||
def angle_between(v1, v2):
|
||||
return np.arccos(np.dot(v1, v2) / np.linalg.norm(v1) / np.linalg.norm(v2))
|
||||
|
||||
|
||||
class Grid(object):
|
||||
|
||||
def __init__(self, graph):
|
||||
|
||||
self._graph = graph
|
||||
self._structure = self._make_structure()
|
||||
self._walls = self._make_walls()
|
||||
|
||||
def search_on_grid(self, x, y):
|
||||
visit = [[0, 1], [0, -1], [1, 0], [1, 1],
|
||||
[1, -1], [-1, 0], [-1, 1], [-1, -1]]
|
||||
c_x, c_y = x, y
|
||||
scale = 1
|
||||
while self._structure[c_x, c_y] != 0:
|
||||
for offset in visit:
|
||||
c_x, c_y = x + offset[0] * scale, y + offset[1] * scale
|
||||
|
||||
if c_x >= 0 and c_x < self._graph.get_resolution()[
|
||||
0] and c_y >= 0 and c_y < self._graph.get_resolution()[1]:
|
||||
if self._structure[c_x, c_y] == 0:
|
||||
break
|
||||
else:
|
||||
c_x, c_y = x, y
|
||||
scale += 1
|
||||
|
||||
return c_x, c_y
|
||||
def get_walls(self):
|
||||
return self._walls
|
||||
|
||||
def get_wall_source(self, pos, pos_ori, target):
|
||||
|
||||
free_nodes = self._get_adjacent_free_nodes(pos)
|
||||
# print self._walls
|
||||
final_walls = copy.copy(self._walls)
|
||||
# print final_walls
|
||||
heading_start = np.array([pos_ori[0], pos_ori[1]])
|
||||
for adj in free_nodes:
|
||||
|
||||
start_to_goal = np.array([adj[0] - pos[0], adj[1] - pos[1]])
|
||||
angle = angle_between(heading_start, start_to_goal)
|
||||
if (angle > 1.6 and adj != target):
|
||||
final_walls.add((adj[0], adj[1]))
|
||||
|
||||
return final_walls
|
||||
|
||||
def get_wall_target(self, pos, pos_ori, source):
|
||||
|
||||
free_nodes = self._get_adjacent_free_nodes(pos)
|
||||
final_walls = copy.copy(self._walls)
|
||||
heading_start = np.array([pos_ori[0], pos_ori[1]])
|
||||
for adj in free_nodes:
|
||||
|
||||
start_to_goal = np.array([adj[0] - pos[0], adj[1] - pos[1]])
|
||||
angle = angle_between(heading_start, start_to_goal)
|
||||
|
||||
if (angle < 1.0 and adj != source):
|
||||
final_walls.add((adj[0], adj[1]))
|
||||
|
||||
return final_walls
|
||||
|
||||
def _draw_line(self, grid, xi, yi, xf, yf):
|
||||
|
||||
if xf < xi:
|
||||
aux = xi
|
||||
xi = xf
|
||||
xf = aux
|
||||
|
||||
if yf < yi:
|
||||
aux = yi
|
||||
yi = yf
|
||||
yf = aux
|
||||
|
||||
for i in range(xi, xf + 1):
|
||||
|
||||
for j in range(yi, yf + 1):
|
||||
grid[i, j] = 0.0
|
||||
|
||||
return grid
|
||||
|
||||
def _make_structure(self):
|
||||
structure = np.ones(
|
||||
(self._graph.get_resolution()[0],
|
||||
self._graph.get_resolution()[1]))
|
||||
|
||||
for key, connections in self._graph.get_edges().items():
|
||||
|
||||
# draw a line
|
||||
for con in connections:
|
||||
# print key[0],key[1],con[0],con[1]
|
||||
structure = self._draw_line(
|
||||
structure, key[0], key[1], con[0], con[1])
|
||||
# print grid
|
||||
return structure
|
||||
|
||||
def _make_walls(self):
|
||||
walls = set()
|
||||
|
||||
for i in range(self._structure.shape[0]):
|
||||
|
||||
for j in range(self._structure.shape[1]):
|
||||
if self._structure[i, j] == 1.0:
|
||||
walls.add((i, j))
|
||||
|
||||
return walls
|
||||
|
||||
def _get_adjacent_free_nodes(self, pos):
|
||||
""" Eight nodes in total """
|
||||
visit = [[0, 1], [0, -1], [1, 0], [1, 1],
|
||||
[1, -1], [-1, 0], [-1, 1], [-1, -1]]
|
||||
|
||||
adjacent = set()
|
||||
for offset in visit:
|
||||
node = (pos[0] + offset[0], pos[1] + offset[1])
|
||||
|
||||
if (node[0] >= 0 and node[0] < self._graph.get_resolution()[0]
|
||||
and node[1] >= 0 and node[1] < self._graph.get_resolution()[1]):
|
||||
|
||||
if self._structure[node[0], node[1]] == 0.0:
|
||||
adjacent.add(node)
|
||||
|
||||
return adjacent
|
||||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -19,54 +19,36 @@ try:
|
|||
except ImportError:
|
||||
raise RuntimeError('cannot import PIL, make sure pillow package is installed')
|
||||
|
||||
|
||||
def string_to_node(string):
|
||||
vec = string.split(',')
|
||||
return (int(vec[0]), int(vec[1]))
|
||||
from carla.planner.graph import Graph
|
||||
from carla.planner.graph import sldist
|
||||
from carla.planner.grid import Grid
|
||||
from carla.planner.converter import Converter
|
||||
|
||||
|
||||
def string_to_floats(string):
|
||||
vec = string.split(',')
|
||||
return (float(vec[0]), float(vec[1]), float(vec[2]))
|
||||
def color_to_angle(color):
|
||||
return (float(color) / 255.0) * 2 * math.pi
|
||||
|
||||
|
||||
class CarlaMap(object):
|
||||
def __init__(self, city):
|
||||
|
||||
def __init__(self, city, pixel_density, node_density):
|
||||
dir_path = os.path.dirname(__file__)
|
||||
city_file = os.path.join(dir_path, city + '.txt')
|
||||
|
||||
city_map_file = os.path.join(dir_path, city + '.png')
|
||||
city_map_file_lanes = os.path.join(dir_path, city + 'Lanes.png')
|
||||
city_map_file_center = os.path.join(dir_path, city + 'Central.png')
|
||||
|
||||
with open(city_file, 'r') as file_object:
|
||||
# The built graph. This is the exact same graph that unreal builds. This
|
||||
# is a generic structure used for many cases
|
||||
self._graph = Graph(city_file, node_density)
|
||||
|
||||
linewordloffset = file_object.readline()
|
||||
# The offset of the world from the zero coordinates ( The
|
||||
# coordinate we consider zero)
|
||||
self.worldoffset = string_to_floats(linewordloffset)
|
||||
self._pixel_density = pixel_density
|
||||
self._grid = Grid(self._graph)
|
||||
# The number of game units per pixel. For now this is fixed.
|
||||
|
||||
lineworldangles = file_object.readline()
|
||||
self.angles = string_to_floats(lineworldangles)
|
||||
self._converter = Converter(city_file, pixel_density, node_density)
|
||||
|
||||
self.worldrotation = np.array([
|
||||
[math.cos(math.radians(self.angles[2])), -math.sin(math.radians(self.angles[2])), 0.0],
|
||||
[math.sin(math.radians(self.angles[2])), math.cos(math.radians(self.angles[2])), 0.0],
|
||||
[0.0, 0.0, 1.0]])
|
||||
|
||||
# Ignore for now, these are offsets for map coordinates and scale
|
||||
# (not used).
|
||||
_ = file_object.readline()
|
||||
linemapoffset = file_object.readline()
|
||||
|
||||
# The offset of the map zero coordinate.
|
||||
self.mapoffset = string_to_floats(linemapoffset)
|
||||
|
||||
# the graph resolution.
|
||||
linegraphres = file_object.readline()
|
||||
self.resolution = string_to_node(linegraphres)
|
||||
|
||||
# The number of game units per pixel.
|
||||
self.pixel_density = 16.43
|
||||
# Load the lanes image
|
||||
self.map_image_lanes = Image.open(city_map_file_lanes)
|
||||
self.map_image_lanes.load()
|
||||
|
|
@ -76,72 +58,95 @@ class CarlaMap(object):
|
|||
self.map_image.load()
|
||||
self.map_image = np.asarray(self.map_image, dtype="int32")
|
||||
|
||||
# Load the lanes image
|
||||
self.map_image_center = Image.open(city_map_file_center)
|
||||
self.map_image_center.load()
|
||||
self.map_image_center = np.asarray(self.map_image_center, dtype="int32")
|
||||
|
||||
def get_graph_resolution(self):
|
||||
|
||||
return self._graph.get_resolution()
|
||||
|
||||
def get_map(self, height=None):
|
||||
if height is not None:
|
||||
img = Image.fromarray(self.map_image.astype(np.uint8))
|
||||
|
||||
aspect_ratio = height/float(self.map_image.shape[0])
|
||||
aspect_ratio = height / float(self.map_image.shape[0])
|
||||
|
||||
img = img.resize((int(aspect_ratio*self.map_image.shape[1]),height), Image.ANTIALIAS)
|
||||
img = img.resize((int(aspect_ratio * self.map_image.shape[1]), height), Image.ANTIALIAS)
|
||||
img.load()
|
||||
return np.asarray(img, dtype="int32")
|
||||
return np.fliplr(self.map_image)
|
||||
|
||||
def get_map_lanes(self, height=None):
|
||||
# if size is not None:
|
||||
# img = Image.fromarray(self.map_image_lanes.astype(np.uint8))
|
||||
# img = img.resize((size[1], size[0]), Image.ANTIALIAS)
|
||||
# img.load()
|
||||
# return np.fliplr(np.asarray(img, dtype="int32"))
|
||||
# return np.fliplr(self.map_image_lanes)
|
||||
raise NotImplementedError
|
||||
|
||||
def get_position_on_map(self, world):
|
||||
"""Get the position on the map for a certain world position."""
|
||||
relative_location = []
|
||||
pixel = []
|
||||
|
||||
rotation = np.array([world[0], world[1], world[2]])
|
||||
rotation = rotation.dot(self.worldrotation)
|
||||
|
||||
relative_location.append(rotation[0] + self.worldoffset[0] - self.mapoffset[0])
|
||||
relative_location.append(rotation[1] + self.worldoffset[1] - self.mapoffset[1])
|
||||
relative_location.append(rotation[2] + self.worldoffset[2] - self.mapoffset[2])
|
||||
|
||||
pixel.append(math.floor(relative_location[0] / float(self.pixel_density)))
|
||||
pixel.append(math.floor(relative_location[1] / float(self.pixel_density)))
|
||||
|
||||
return pixel
|
||||
|
||||
def get_position_on_world(self, pixel):
|
||||
"""Get world position of a certain map position."""
|
||||
relative_location = []
|
||||
world_vertex = []
|
||||
relative_location.append(pixel[0] * self.pixel_density)
|
||||
relative_location.append(pixel[1] * self.pixel_density)
|
||||
|
||||
world_vertex.append(relative_location[0] + self.mapoffset[0] - self.worldoffset[0])
|
||||
world_vertex.append(relative_location[1] + self.mapoffset[1] - self.worldoffset[1])
|
||||
world_vertex.append(22) # Z does not matter for now.
|
||||
|
||||
return world_vertex
|
||||
def get_map_lanes(self, size=None):
|
||||
if size is not None:
|
||||
img = Image.fromarray(self.map_image_lanes.astype(np.uint8))
|
||||
img = img.resize((size[1], size[0]), Image.ANTIALIAS)
|
||||
img.load()
|
||||
return np.fliplr(np.asarray(img, dtype="int32"))
|
||||
return np.fliplr(self.map_image_lanes)
|
||||
|
||||
def get_lane_orientation(self, world):
|
||||
"""Get the lane orientation of a certain world position."""
|
||||
relative_location = []
|
||||
pixel = []
|
||||
rotation = np.array([world[0], world[1], world[2]])
|
||||
rotation = rotation.dot(self.worldrotation)
|
||||
|
||||
relative_location.append(rotation[0] + self.worldoffset[0] - self.mapoffset[0])
|
||||
relative_location.append(rotation[1] + self.worldoffset[1] - self.mapoffset[1])
|
||||
relative_location.append(rotation[2] + self.worldoffset[2] - self.mapoffset[2])
|
||||
|
||||
pixel.append(math.floor(relative_location[0] / float(self.pixel_density)))
|
||||
pixel.append(math.floor(relative_location[1] / float(self.pixel_density)))
|
||||
pixel = self.convert_to_pixel(world)
|
||||
|
||||
ori = self.map_image_lanes[int(pixel[1]), int(pixel[0]), 2]
|
||||
ori = ((float(ori) / 255.0)) * 2 * math.pi
|
||||
ori = color_to_angle(ori)
|
||||
|
||||
return (-math.cos(ori), -math.sin(ori))
|
||||
|
||||
def convert_to_node(self, input_data):
|
||||
"""
|
||||
Receives a data type (Can Be Pixel or World )
|
||||
:param input_data: position in some coordinate
|
||||
:return: A node object
|
||||
"""
|
||||
return self._converter.convert_to_node(input_data)
|
||||
|
||||
def convert_to_pixel(self, input_data):
|
||||
"""
|
||||
Receives a data type (Can Be Node or World )
|
||||
:param input_data: position in some coordinate
|
||||
:return: A node object
|
||||
"""
|
||||
return self._converter.convert_to_pixel(input_data)
|
||||
|
||||
def convert_to_world(self, input_data):
|
||||
"""
|
||||
Receives a data type (Can Be Pixel or Node )
|
||||
:param input_data: position in some coordinate
|
||||
:return: A node object
|
||||
"""
|
||||
return self._converter.convert_to_world(input_data)
|
||||
|
||||
def get_walls_directed(self, node_source, source_ori, node_target, target_ori):
|
||||
"""
|
||||
This is the most hacky function. Instead of planning on two ways,
|
||||
we basically use a one way road and interrupt the other road by adding
|
||||
an artificial wall.
|
||||
|
||||
"""
|
||||
|
||||
final_walls = self._grid.get_wall_source(node_source, source_ori, node_target)
|
||||
|
||||
final_walls = final_walls.union(self._grid.get_wall_target(
|
||||
node_target, target_ori, node_source))
|
||||
return final_walls
|
||||
|
||||
def get_walls(self):
|
||||
|
||||
return self._grid.get_walls()
|
||||
|
||||
def get_distance_closest_node(self, pos):
|
||||
|
||||
distance = []
|
||||
for node_iter in self._graph.intersection_nodes():
|
||||
distance.append(sldist(node_iter, pos))
|
||||
|
||||
return sorted(distance)[0]
|
||||
|
||||
def get_intersection_nodes(self):
|
||||
return self._graph.intersection_nodes()
|
||||
|
||||
def search_on_grid(self,node):
|
||||
return self._grid.search_on_grid(node[0], node[1])
|
||||
|
|
|
|||
|
|
@ -0,0 +1,173 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import collections
|
||||
import math
|
||||
|
||||
import numpy as np
|
||||
|
||||
from . import city_track
|
||||
|
||||
|
||||
def compare(x, y):
|
||||
return collections.Counter(x) == collections.Counter(y)
|
||||
|
||||
|
||||
|
||||
# Constants Used for the high level commands
|
||||
|
||||
|
||||
REACH_GOAL = 0.0
|
||||
GO_STRAIGHT = 5.0
|
||||
TURN_RIGHT = 4.0
|
||||
TURN_LEFT = 3.0
|
||||
LANE_FOLLOW = 2.0
|
||||
|
||||
|
||||
# Auxiliary algebra function
|
||||
def angle_between(v1, v2):
|
||||
return np.arccos(np.dot(v1, v2) / np.linalg.norm(v1) / np.linalg.norm(v2))
|
||||
|
||||
|
||||
def sldist(c1, c2): return math.sqrt((c2[0] - c1[0]) ** 2 + (c2[1] - c1[1]) ** 2)
|
||||
|
||||
|
||||
def signal(v1, v2):
|
||||
return np.cross(v1, v2) / np.linalg.norm(v1) / np.linalg.norm(v2)
|
||||
|
||||
|
||||
class Planner(object):
|
||||
|
||||
def __init__(self, city_name):
|
||||
|
||||
self._city_track = city_track.CityTrack(city_name)
|
||||
|
||||
self._commands = []
|
||||
|
||||
def get_next_command(self, source, source_ori, target, target_ori):
|
||||
"""
|
||||
Computes the full plan and returns the next command,
|
||||
:param source: source position
|
||||
:param source_ori: source orientation
|
||||
:param target: target position
|
||||
:param target_ori: target orientation
|
||||
:return: a command ( Straight,Lane Follow, Left or Right)
|
||||
"""
|
||||
|
||||
track_source = self._city_track.project_node(source)
|
||||
track_target = self._city_track.project_node(target)
|
||||
|
||||
# reach the goal
|
||||
|
||||
if self._city_track.is_at_goal(track_source, track_target):
|
||||
return REACH_GOAL
|
||||
|
||||
if (self._city_track.is_at_new_node(track_source)
|
||||
and self._city_track.is_away_from_intersection(track_source)):
|
||||
|
||||
route = self._city_track.compute_route(track_source, source_ori,
|
||||
track_target, target_ori)
|
||||
if route is None:
|
||||
raise RuntimeError('Impossible to find route')
|
||||
|
||||
self._commands = self._route_to_commands(route)
|
||||
|
||||
if self._city_track.is_far_away_from_route_intersection(
|
||||
track_source):
|
||||
return LANE_FOLLOW
|
||||
else:
|
||||
if self._commands:
|
||||
return self._commands[0]
|
||||
else:
|
||||
return LANE_FOLLOW
|
||||
else:
|
||||
|
||||
if self._city_track.is_far_away_from_route_intersection(
|
||||
track_source):
|
||||
return LANE_FOLLOW
|
||||
|
||||
# If there is computed commands
|
||||
if self._commands:
|
||||
return self._commands[0]
|
||||
else:
|
||||
return LANE_FOLLOW
|
||||
|
||||
def get_shortest_path_distance(
|
||||
self,
|
||||
source,
|
||||
source_ori,
|
||||
target,
|
||||
target_ori):
|
||||
|
||||
distance = 0
|
||||
track_source = self._city_track.project_node(source)
|
||||
track_target = self._city_track.project_node(target)
|
||||
|
||||
current_pos = track_source
|
||||
|
||||
route = self._city_track.compute_route(track_source, source_ori,
|
||||
track_target, target_ori)
|
||||
# No Route, distance is zero
|
||||
if route is None:
|
||||
return 0.0
|
||||
|
||||
for node_iter in route:
|
||||
distance += sldist(node_iter, current_pos)
|
||||
current_pos = node_iter
|
||||
|
||||
# We multiply by these values to convert distance to world coordinates
|
||||
return distance * self._city_track.get_pixel_density() \
|
||||
* self._city_track.get_node_density()
|
||||
|
||||
def is_there_posible_route(self, source, source_ori, target, target_ori):
|
||||
|
||||
track_source = self._city_track.project_node(source)
|
||||
track_target = self._city_track.project_node(target)
|
||||
|
||||
return not self._city_track.compute_route(
|
||||
track_source, source_ori, track_target, target_ori) is None
|
||||
|
||||
def test_position(self, source):
|
||||
|
||||
node_source = self._city_track.project_node(source)
|
||||
|
||||
return self._city_track.is_away_from_intersection(node_source)
|
||||
|
||||
def _route_to_commands(self, route):
|
||||
|
||||
"""
|
||||
from the shortest path graph, transform it into a list of commands
|
||||
|
||||
:param route: the sub graph containing the shortest path
|
||||
:return: list of commands encoded from 0-5
|
||||
"""
|
||||
|
||||
commands_list = []
|
||||
|
||||
for i in range(0, len(route)):
|
||||
if route[i] not in self._city_track.get_intersection_nodes():
|
||||
continue
|
||||
|
||||
current = route[i]
|
||||
past = route[i - 1]
|
||||
future = route[i + 1]
|
||||
|
||||
past_to_current = np.array(
|
||||
[current[0] - past[0], current[1] - past[1]])
|
||||
current_to_future = np.array(
|
||||
[future[0] - current[0], future[1] - current[1]])
|
||||
angle = signal(current_to_future, past_to_current)
|
||||
|
||||
if angle < -0.1:
|
||||
command = TURN_RIGHT
|
||||
elif angle > 0.1:
|
||||
command = TURN_LEFT
|
||||
else:
|
||||
command = GO_STRAIGHT
|
||||
|
||||
commands_list.append(command)
|
||||
|
||||
return commands_list
|
||||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -8,6 +8,8 @@
|
|||
|
||||
|
||||
import os
|
||||
import numpy as np
|
||||
import json
|
||||
|
||||
try:
|
||||
import numpy
|
||||
|
|
@ -39,7 +41,12 @@ class Sensor(object):
|
|||
"""
|
||||
Base class for sensor descriptions. Used to add sensors to CarlaSettings.
|
||||
"""
|
||||
pass
|
||||
|
||||
def set(self, **kwargs):
|
||||
for key, value in kwargs.items():
|
||||
if not hasattr(self, key):
|
||||
raise ValueError('CarlaSettings.Camera: no key named %r' % key)
|
||||
setattr(self, key, value)
|
||||
|
||||
|
||||
class Camera(Sensor):
|
||||
|
|
@ -62,12 +69,6 @@ class Camera(Sensor):
|
|||
self.CameraRotationYaw = 0
|
||||
self.set(**kwargs)
|
||||
|
||||
def set(self, **kwargs):
|
||||
for key, value in kwargs.items():
|
||||
if not hasattr(self, key):
|
||||
raise ValueError('CarlaSettings.Camera: no key named %r' % key)
|
||||
setattr(self, key, value)
|
||||
|
||||
def set_image_size(self, pixels_x, pixels_y):
|
||||
self.ImageSizeX = pixels_x
|
||||
self.ImageSizeY = pixels_y
|
||||
|
|
@ -83,6 +84,51 @@ class Camera(Sensor):
|
|||
self.CameraRotationYaw = yaw
|
||||
|
||||
|
||||
class Lidar(Sensor):
|
||||
"""
|
||||
Lidar description. This class can be added to a CarlaSettings object to add
|
||||
a Lidar to the player vehicle.
|
||||
"""
|
||||
|
||||
def __init__(self, name, **kwargs):
|
||||
self.LidarName = name
|
||||
# Number of lasers
|
||||
self.Channels = 32
|
||||
# Measure distance
|
||||
self.Range = 5000
|
||||
# Points generated by all lasers per second
|
||||
self.PointsPerSecond = 100000
|
||||
# Lidar rotation frequency
|
||||
self.RotationFrequency = 10
|
||||
# Upper laser angle, counts from horizontal,
|
||||
# positive values means above horizontal line
|
||||
self.UpperFovLimit = 10
|
||||
# Lower laser angle, counts from horizontal,
|
||||
# negative values means under horizontal line
|
||||
self.LowerFovLimit = -30
|
||||
# wether to show debug points of laser hits in simulator
|
||||
self.ShowDebugPoints = False
|
||||
# Position relative to the player.
|
||||
self.LidarPositionX = 0
|
||||
self.LidarPositionY = 0
|
||||
self.LidarPositionZ = 250
|
||||
# Rotation relative to the player.
|
||||
self.LidarRotationPitch = 0
|
||||
self.LidarRotationRoll = 0
|
||||
self.LidarRotationYaw = 0
|
||||
self.set(**kwargs)
|
||||
|
||||
def set_position(self, x, y, z):
|
||||
self.LidarPositionX = x
|
||||
self.LidarPositionY = y
|
||||
self.LidarPositionZ = z
|
||||
|
||||
def set_rotation(self, pitch, roll, yaw):
|
||||
self.LidarRotationPitch = pitch
|
||||
self.LidarRotationRoll = roll
|
||||
self.LidarRotationYaw = yaw
|
||||
|
||||
|
||||
# ==============================================================================
|
||||
# -- SensorData ----------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
|
@ -252,3 +298,62 @@ class PointCloud(SensorData):
|
|||
|
||||
def __str__(self):
|
||||
return str(self.array)
|
||||
|
||||
|
||||
class LidarMeasurement(SensorData):
|
||||
"""Data generated by a Lidar."""
|
||||
|
||||
def __init__(self, horizontal_angle, channels_count, lidar_type, raw_data):
|
||||
self.horizontal_angle = horizontal_angle
|
||||
self.channels_count = channels_count
|
||||
self.type = lidar_type
|
||||
self._converted_data = None
|
||||
|
||||
points_count_by_channel_size = channels_count * 4
|
||||
points_count_by_channel_bytes = raw_data[4*4:4*4 + points_count_by_channel_size]
|
||||
self.points_count_by_channel = np.frombuffer(points_count_by_channel_bytes, dtype=np.dtype('uint32'))
|
||||
|
||||
self.points_size = int(np.sum(self.points_count_by_channel) * 3 * 8) # three floats X, Y, Z
|
||||
begin = 4*4 + points_count_by_channel_size # 4*4 is horizontal_angle, type, channels_count
|
||||
end = begin + self.points_size
|
||||
self.points_data = raw_data[begin:end]
|
||||
|
||||
self._data_size = 4*4 + points_count_by_channel_size + self.points_size
|
||||
|
||||
@property
|
||||
def size_in_bytes(self):
|
||||
return self._data_size
|
||||
|
||||
@property
|
||||
def data(self):
|
||||
"""
|
||||
Lazy initialization for data property, stores converted data in its
|
||||
default format.
|
||||
"""
|
||||
if self._converted_data is None:
|
||||
|
||||
points_in_one_channel = self.points_count_by_channel[0]
|
||||
points = np.frombuffer(self.points_data[:self.points_size], dtype='float')
|
||||
points = np.reshape(points, (self.channels_count, points_in_one_channel, 3))
|
||||
|
||||
self._converted_data = {
|
||||
'horizontal_angle' : self.horizontal_angle,
|
||||
'channels_count' : self.channels_count,
|
||||
'points_count_by_channel' : self.points_count_by_channel,
|
||||
'points' : points
|
||||
}
|
||||
|
||||
return self._converted_data
|
||||
|
||||
def save_to_disk(self, filename):
|
||||
"""Save lidar measurements to disk"""
|
||||
folder = os.path.dirname(filename)
|
||||
if not os.path.isdir(folder):
|
||||
os.makedirs(folder)
|
||||
with open(filename, 'wt') as f:
|
||||
f.write(json.dumps({
|
||||
'horizontal_angle' : self.horizontal_angle,
|
||||
'channels_count' : self.channels_count,
|
||||
'points_count_by_channel' : self.points_count_by_channel.tolist(),
|
||||
'points' : self.data['points'].tolist()
|
||||
}))
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -46,6 +46,7 @@ class CarlaSettings(object):
|
|||
self.randomize_weather()
|
||||
self.set(**kwargs)
|
||||
self._cameras = []
|
||||
self._lidars = []
|
||||
|
||||
def set(self, **kwargs):
|
||||
for key, value in kwargs.items():
|
||||
|
|
@ -69,13 +70,15 @@ class CarlaSettings(object):
|
|||
"""Add a sensor to the player vehicle (see sensor.py)."""
|
||||
if isinstance(sensor, carla_sensor.Camera):
|
||||
self._cameras.append(sensor)
|
||||
elif isinstance(sensor, carla_sensor.Lidar):
|
||||
self._lidars.append(sensor)
|
||||
else:
|
||||
raise ValueError('Sensor not supported')
|
||||
|
||||
def __str__(self):
|
||||
"""Converts this object to an INI formatted string."""
|
||||
ini = ConfigParser()
|
||||
ini.optionxform=str
|
||||
ini.optionxform = str
|
||||
S_SERVER = 'CARLA/Server'
|
||||
S_LEVEL = 'CARLA/LevelSettings'
|
||||
S_CAPTURE = 'CARLA/SceneCapture'
|
||||
|
|
@ -99,6 +102,7 @@ class CarlaSettings(object):
|
|||
|
||||
ini.add_section(S_CAPTURE)
|
||||
ini.set(S_CAPTURE, 'Cameras', ','.join(c.CameraName for c in self._cameras))
|
||||
ini.set(S_CAPTURE, 'Lidars', ','.join(l.LidarName for l in self._lidars))
|
||||
|
||||
for camera in self._cameras:
|
||||
add_section(S_CAPTURE + '/' + camera.CameraName, camera, [
|
||||
|
|
@ -113,6 +117,22 @@ class CarlaSettings(object):
|
|||
'CameraRotationRoll',
|
||||
'CameraRotationYaw'])
|
||||
|
||||
for lidar in self._lidars:
|
||||
add_section(S_CAPTURE + '/' + lidar.LidarName, lidar, [
|
||||
'Channels',
|
||||
'Range',
|
||||
'PointsPerSecond',
|
||||
'RotationFrequency',
|
||||
'UpperFovLimit',
|
||||
'LowerFovLimit',
|
||||
'ShowDebugPoints',
|
||||
'LidarPositionX',
|
||||
'LidarPositionY',
|
||||
'LidarPositionZ',
|
||||
'LidarRotationPitch',
|
||||
'LidarRotationRoll',
|
||||
'LidarRotationYaw'])
|
||||
|
||||
if sys.version_info >= (3, 0):
|
||||
text = io.StringIO()
|
||||
else:
|
||||
|
|
@ -129,14 +149,15 @@ def get_sensor_names(settings):
|
|||
"""
|
||||
if isinstance(settings, CarlaSettings):
|
||||
# pylint: disable=protected-access
|
||||
return [camera.CameraName for camera in settings._cameras]
|
||||
return [camera.CameraName for camera in settings._cameras] + \
|
||||
[lidar.LidarName for lidar in settings._lidars]
|
||||
ini = ConfigParser()
|
||||
if sys.version_info >= (3, 2):
|
||||
ini.read_string(settings)
|
||||
elif sys.version_info >= (3, 0):
|
||||
ini.readfp(io.StringIO(settings)) # pylint: disable=deprecated-method
|
||||
ini.readfp(io.StringIO(settings)) # pylint: disable=deprecated-method
|
||||
else:
|
||||
ini.readfp(io.BytesIO(settings)) # pylint: disable=deprecated-method
|
||||
ini.readfp(io.BytesIO(settings)) # pylint: disable=deprecated-method
|
||||
|
||||
section_name = 'CARLA/SceneCapture'
|
||||
option_name = 'Cameras'
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -41,7 +41,7 @@ class TCPClient(object):
|
|||
self._socket.settimeout(self._timeout)
|
||||
logging.debug('%sconnected', self._logprefix)
|
||||
return
|
||||
except OSError as exception:
|
||||
except socket.error as exception:
|
||||
error = exception
|
||||
logging.debug('%sconnection attempt %d: %s', self._logprefix, attempt, error)
|
||||
time.sleep(1)
|
||||
|
|
@ -65,7 +65,7 @@ class TCPClient(object):
|
|||
header = struct.pack('<L', len(message))
|
||||
try:
|
||||
self._socket.sendall(header + message)
|
||||
except OSError as exception:
|
||||
except socket.error as exception:
|
||||
self._reraise_exception_as_tcp_error('failed to write data', exception)
|
||||
|
||||
def read(self):
|
||||
|
|
@ -85,7 +85,7 @@ class TCPClient(object):
|
|||
while length > 0:
|
||||
try:
|
||||
data = self._socket.recv(length)
|
||||
except OSError as exception:
|
||||
except socket.error as exception:
|
||||
self._reraise_exception_as_tcp_error('failed to read data', exception)
|
||||
if not data:
|
||||
raise TCPConnectionError(self._logprefix + 'connection closed')
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -16,13 +16,13 @@ import random
|
|||
import time
|
||||
|
||||
from carla.client import make_carla_client
|
||||
from carla.sensor import Camera
|
||||
from carla.sensor import Camera, Lidar, LidarMeasurement
|
||||
from carla.settings import CarlaSettings
|
||||
from carla.tcp import TCPConnectionError
|
||||
from carla.util import print_over_same_line
|
||||
|
||||
|
||||
def run_carla_client(host, port, autopilot_on, save_images_to_disk, image_filename_format, settings_filepath):
|
||||
def run_carla_client(host, port, autopilot_on, save_images_to_disk, image_filename_format, lidar_filename_format, settings_filepath):
|
||||
# Here we will run 3 episodes with 300 frames each.
|
||||
number_of_episodes = 3
|
||||
frames_per_episode = 300
|
||||
|
|
@ -70,6 +70,19 @@ def run_carla_client(host, port, autopilot_on, save_images_to_disk, image_filena
|
|||
camera1.set_position(30, 0, 130)
|
||||
settings.add_sensor(camera1)
|
||||
|
||||
lidar0 = Lidar('Lidar32')
|
||||
lidar0.set_position(0, 0, 250)
|
||||
lidar0.set_rotation(0, 0, 0)
|
||||
lidar0.set(
|
||||
Channels = 32,
|
||||
Range = 5000,
|
||||
PointsPerSecond = 100000,
|
||||
RotationFrequency = 10,
|
||||
UpperFovLimit = 10,
|
||||
LowerFovLimit = -30,
|
||||
ShowDebugPoints = False)
|
||||
settings.add_sensor(lidar0)
|
||||
|
||||
else:
|
||||
|
||||
# Alternatively, we can load these settings from a file.
|
||||
|
|
@ -103,8 +116,11 @@ def run_carla_client(host, port, autopilot_on, save_images_to_disk, image_filena
|
|||
|
||||
# Save the images to disk if requested.
|
||||
if save_images_to_disk:
|
||||
for name, image in sensor_data.items():
|
||||
image.save_to_disk(image_filename_format.format(episode, name, frame))
|
||||
for name, measurement in sensor_data.items():
|
||||
if isinstance(measurement, LidarMeasurement):
|
||||
measurement.save_to_disk(lidar_filename_format.format(episode, name, frame))
|
||||
else:
|
||||
measurement.save_to_disk(image_filename_format.format(episode, name, frame))
|
||||
|
||||
# We can access the encoded data of a given image as numpy
|
||||
# array using its "data" property. For instance, to get the
|
||||
|
|
@ -209,9 +225,11 @@ def main():
|
|||
autopilot_on=args.autopilot,
|
||||
save_images_to_disk=args.images_to_disk,
|
||||
image_filename_format='_out/episode_{:0>4d}/{:s}/{:0>6d}.png',
|
||||
lidar_filename_format='_out/episode_{:0>4d}/{:s}/{:0>6d}.json',
|
||||
settings_filepath=args.carla_settings)
|
||||
|
||||
print('Done.')
|
||||
return
|
||||
|
||||
except TCPConnectionError as error:
|
||||
logging.error(error)
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -93,6 +93,19 @@ def make_carla_settings():
|
|||
camera2.set_position(200, 0, 140)
|
||||
camera2.set_rotation(0.0, 0.0, 0.0)
|
||||
settings.add_sensor(camera2)
|
||||
lidar0 = sensor.Lidar('Lidar32')
|
||||
lidar0.set_position(0, 0, 250)
|
||||
lidar0.set_rotation(0, 0, 0)
|
||||
lidar0.set(
|
||||
Channels = 32,
|
||||
Range = 5000,
|
||||
PointsPerSecond = 100000,
|
||||
RotationFrequency = 10,
|
||||
UpperFovLimit = 10,
|
||||
LowerFovLimit = -30,
|
||||
ShowDebugPoints = False)
|
||||
settings.add_sensor(lidar0)
|
||||
|
||||
return settings
|
||||
|
||||
|
||||
|
|
@ -124,10 +137,11 @@ class CarlaGame(object):
|
|||
self._main_image = None
|
||||
self._mini_view_image1 = None
|
||||
self._mini_view_image2 = None
|
||||
self._lidar_measurement = None
|
||||
self._map_view = None
|
||||
self._is_on_reverse = False
|
||||
self._city_name = city_name
|
||||
self._map = CarlaMap(city_name) if city_name is not None else None
|
||||
self._map = CarlaMap(city_name, 16.43, 50.0) if city_name is not None else None
|
||||
self._map_shape = self._map.map_image.shape if city_name is not None else None
|
||||
self._map_view = self._map.get_map(WINDOW_HEIGHT) if city_name is not None else None
|
||||
self._position = None
|
||||
|
|
@ -177,12 +191,13 @@ class CarlaGame(object):
|
|||
self._main_image = sensor_data['CameraRGB']
|
||||
self._mini_view_image1 = sensor_data['CameraDepth']
|
||||
self._mini_view_image2 = sensor_data['CameraSemSeg']
|
||||
self._lidar_measurement = sensor_data['Lidar32']
|
||||
|
||||
# Print measurements every second.
|
||||
if self._timer.elapsed_seconds_since_lap() > 1.0:
|
||||
if self._city_name is not None:
|
||||
# Function to get car position on map.
|
||||
map_position = self._map.get_position_on_map([
|
||||
map_position = self._map.convert_to_pixel([
|
||||
measurements.player_measurements.transform.location.x,
|
||||
measurements.player_measurements.transform.location.y,
|
||||
measurements.player_measurements.transform.location.z])
|
||||
|
|
@ -206,7 +221,7 @@ class CarlaGame(object):
|
|||
control = self._get_keyboard_control(pygame.key.get_pressed())
|
||||
# Set the player position
|
||||
if self._city_name is not None:
|
||||
self._position = self._map.get_position_on_map([
|
||||
self._position = self._map.convert_to_pixel([
|
||||
measurements.player_measurements.transform.location.x,
|
||||
measurements.player_measurements.transform.location.y,
|
||||
measurements.player_measurements.transform.location.z])
|
||||
|
|
@ -295,25 +310,45 @@ class CarlaGame(object):
|
|||
self._display.blit(
|
||||
surface, (2 * gap_x + MINI_WINDOW_WIDTH, mini_image_y))
|
||||
|
||||
if self._lidar_measurement is not None:
|
||||
|
||||
lidar_data = np.array(self._lidar_measurement.data['points'][:, :, :2])
|
||||
lidar_data /= 50.0
|
||||
lidar_data += 100.0
|
||||
lidar_data = np.fabs(lidar_data)
|
||||
lidar_data = lidar_data.astype(np.int32)
|
||||
lidar_data = np.reshape(lidar_data, (-1, 2))
|
||||
#draw lidar
|
||||
lidar_img_size = (200, 200, 3)
|
||||
lidar_img = np.zeros(lidar_img_size)
|
||||
lidar_img[tuple(lidar_data.T)] = (255, 255, 255)
|
||||
surface = pygame.surfarray.make_surface(
|
||||
lidar_img
|
||||
)
|
||||
self._display.blit(surface, (10, 10))
|
||||
|
||||
if self._map_view is not None:
|
||||
array = self._map_view
|
||||
array = array[:, :, :3]
|
||||
|
||||
new_window_width =(float(WINDOW_HEIGHT)/float(self._map_shape[0]))*float(self._map_shape[1])
|
||||
surface = pygame.surfarray.make_surface(array.swapaxes(0, 1))
|
||||
|
||||
w_pos = int(self._position[0]*(float(WINDOW_HEIGHT)/float(self._map_shape[0])))
|
||||
h_pos =int(self._position[1] *(new_window_width/float(self._map_shape[1])))
|
||||
h_pos = int(self._position[1] *(new_window_width/float(self._map_shape[1])))
|
||||
|
||||
pygame.draw.circle(surface, [255, 0, 0, 255], (w_pos,h_pos), 6, 0)
|
||||
for agent in self._agent_positions:
|
||||
if agent.HasField('vehicle'):
|
||||
agent_position = self._map.get_position_on_map([
|
||||
agent_position = self._map.convert_to_pixel([
|
||||
agent.vehicle.transform.location.x,
|
||||
agent.vehicle.transform.location.y,
|
||||
agent.vehicle.transform.location.z])
|
||||
|
||||
w_pos = int(agent_position[0]*(float(WINDOW_HEIGHT)/float(self._map_shape[0])))
|
||||
h_pos =int(agent_position[1] *(new_window_width/float(self._map_shape[1])))
|
||||
pygame.draw.circle(surface, [255, 0, 255, 255], (w_pos,h_pos), 4, 0)
|
||||
h_pos = int(agent_position[1] *(new_window_width/float(self._map_shape[1])))
|
||||
|
||||
pygame.draw.circle(surface, [255, 0, 255, 255], (w_pos ,h_pos), 4, 0)
|
||||
|
||||
self._display.blit(surface, (WINDOW_WIDTH, 0))
|
||||
|
||||
|
|
|
|||
|
|
@ -2,3 +2,5 @@ Pillow
|
|||
numpy
|
||||
protobuf
|
||||
pygame
|
||||
matplotlib
|
||||
future
|
||||
|
|
|
|||
|
|
@ -0,0 +1,94 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
import argparse
|
||||
import logging
|
||||
import time
|
||||
|
||||
from carla.benchmarks.agent import Agent
|
||||
from carla.benchmarks.corl_2017 import CoRL2017
|
||||
|
||||
from carla.client import make_carla_client, VehicleControl
|
||||
from carla.tcp import TCPConnectionError
|
||||
|
||||
|
||||
class Manual(Agent):
|
||||
"""
|
||||
Sample redefinition of the Agent,
|
||||
An agent that goes straight
|
||||
"""
|
||||
def run_step(self, measurements, sensor_data, target):
|
||||
control = VehicleControl()
|
||||
control.throttle = 0.9
|
||||
|
||||
return control
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
|
||||
argparser = argparse.ArgumentParser(description=__doc__)
|
||||
argparser.add_argument(
|
||||
'-v', '--verbose',
|
||||
action='store_true',
|
||||
dest='verbose',
|
||||
help='print some extra status information')
|
||||
argparser.add_argument(
|
||||
'-db', '--debug',
|
||||
action='store_true',
|
||||
dest='debug',
|
||||
help='print debug information')
|
||||
argparser.add_argument(
|
||||
'--host',
|
||||
metavar='H',
|
||||
default='localhost',
|
||||
help='IP of the host server (default: localhost)')
|
||||
argparser.add_argument(
|
||||
'-p', '--port',
|
||||
metavar='P',
|
||||
default=2000,
|
||||
type=int,
|
||||
help='TCP port to listen to (default: 2000)')
|
||||
argparser.add_argument(
|
||||
'-c', '--city-name',
|
||||
metavar='C',
|
||||
default='Town01',
|
||||
help='The town that is going to be used on benchmark'
|
||||
+ '(needs to match active town in server, options: Town01 or Town02)')
|
||||
argparser.add_argument(
|
||||
'-n', '--log_name',
|
||||
metavar='T',
|
||||
default='test',
|
||||
help='The name of the log file to be created by the benchmark'
|
||||
)
|
||||
|
||||
args = argparser.parse_args()
|
||||
if args.debug:
|
||||
log_level = logging.DEBUG
|
||||
elif args.verbose:
|
||||
log_level = logging.INFO
|
||||
else:
|
||||
log_level = logging.WARNING
|
||||
|
||||
logging.basicConfig(format='%(levelname)s: %(message)s', level=log_level)
|
||||
|
||||
logging.info('listening to server %s:%s', args.host, args.port)
|
||||
|
||||
while True:
|
||||
try:
|
||||
with make_carla_client(args.host, args.port) as client:
|
||||
corl = CoRL2017(city_name=args.city_name, name_to_save=args.log_name)
|
||||
agent = Manual(args.city_name)
|
||||
results = corl.benchmark_agent(agent, client)
|
||||
corl.plot_summary_test()
|
||||
corl.plot_summary_train()
|
||||
|
||||
break
|
||||
|
||||
except TCPConnectionError as error:
|
||||
logging.error(error)
|
||||
time.sleep(1)
|
||||
|
|
@ -0,0 +1,15 @@
|
|||
from setuptools import setup
|
||||
|
||||
# @todo Dependencies are missing.
|
||||
|
||||
setup(
|
||||
name='carla_client',
|
||||
version='0.7.1',
|
||||
packages=['carla', 'carla.benchmarks', 'carla.planner'],
|
||||
license='MIT License',
|
||||
description='Python API for communicating with the CARLA server.',
|
||||
url='https://github.com/carla-simulator/carla',
|
||||
author='The CARLA team',
|
||||
author_email='carla.simulator@gmail.com',
|
||||
include_package_data=True
|
||||
)
|
||||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -0,0 +1,168 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
||||
"""Client that runs two servers simultaneously to test repeatability."""
|
||||
|
||||
import argparse
|
||||
import logging
|
||||
import os
|
||||
import random
|
||||
import sys
|
||||
import time
|
||||
|
||||
sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
|
||||
|
||||
from carla.client import make_carla_client
|
||||
from carla.sensor import Camera, Image
|
||||
from carla.settings import CarlaSettings
|
||||
from carla.tcp import TCPConnectionError
|
||||
|
||||
|
||||
def run_carla_clients(args):
|
||||
filename = '_images_repeatability/server{:d}/{:0>6d}.png'
|
||||
with make_carla_client(args.host1, args.port1) as client1:
|
||||
logging.info('1st client connected')
|
||||
with make_carla_client(args.host2, args.port2) as client2:
|
||||
logging.info('2nd client connected')
|
||||
|
||||
settings = CarlaSettings()
|
||||
settings.set(
|
||||
SynchronousMode=True,
|
||||
SendNonPlayerAgentsInfo=True,
|
||||
NumberOfVehicles=50,
|
||||
NumberOfPedestrians=50,
|
||||
WeatherId=random.choice([1, 3, 7, 8, 14]))
|
||||
settings.randomize_seeds()
|
||||
|
||||
if args.images_to_disk:
|
||||
camera = Camera('DefaultCamera')
|
||||
camera.set_image_size(800, 600)
|
||||
settings.add_sensor(camera)
|
||||
|
||||
scene1 = client1.load_settings(settings)
|
||||
scene2 = client2.load_settings(settings)
|
||||
|
||||
number_of_player_starts = len(scene1.player_start_spots)
|
||||
assert number_of_player_starts == len(scene2.player_start_spots)
|
||||
player_start = random.randint(0, max(0, number_of_player_starts - 1))
|
||||
logging.info(
|
||||
'start episode at %d/%d player start (run forever, press ctrl+c to cancel)',
|
||||
player_start,
|
||||
number_of_player_starts)
|
||||
|
||||
client1.start_episode(player_start)
|
||||
client2.start_episode(player_start)
|
||||
|
||||
frame = 0
|
||||
while True:
|
||||
frame += 1
|
||||
|
||||
meas1, sensor_data1 = client1.read_data()
|
||||
meas2, sensor_data2 = client2.read_data()
|
||||
|
||||
player1 = meas1.player_measurements
|
||||
player2 = meas2.player_measurements
|
||||
|
||||
images1 = [x for x in sensor_data1.values() if isinstance(x, Image)]
|
||||
images2 = [x for x in sensor_data2.values() if isinstance(x, Image)]
|
||||
|
||||
control1 = player1.autopilot_control
|
||||
control2 = player2.autopilot_control
|
||||
|
||||
try:
|
||||
assert len(images1) == len(images2)
|
||||
assert len(meas1.non_player_agents) == len(meas2.non_player_agents)
|
||||
assert player1.transform.location.x == player2.transform.location.x
|
||||
assert player1.transform.location.y == player2.transform.location.y
|
||||
assert player1.transform.location.z == player2.transform.location.z
|
||||
assert control1.steer == control2.steer
|
||||
assert control1.throttle == control2.throttle
|
||||
assert control1.brake == control2.brake
|
||||
assert control1.hand_brake == control2.hand_brake
|
||||
assert control1.reverse == control2.reverse
|
||||
except AssertionError:
|
||||
logging.exception('assertion failed')
|
||||
|
||||
if args.images_to_disk:
|
||||
assert len(images1) == 1
|
||||
images1[0].save_to_disk(filename.format(1, frame))
|
||||
images2[0].save_to_disk(filename.format(2, frame))
|
||||
|
||||
client1.send_control(control1)
|
||||
client2.send_control(control2)
|
||||
|
||||
|
||||
def main():
|
||||
argparser = argparse.ArgumentParser(description=__doc__)
|
||||
argparser.add_argument(
|
||||
'-v', '--verbose',
|
||||
action='store_true',
|
||||
dest='debug',
|
||||
help='print debug information')
|
||||
argparser.add_argument(
|
||||
'--log',
|
||||
metavar='LOG_FILE',
|
||||
default=None,
|
||||
help='print output to file')
|
||||
argparser.add_argument(
|
||||
'--host1',
|
||||
metavar='H',
|
||||
default='127.0.0.1',
|
||||
help='IP of the first host server (default: 127.0.0.1)')
|
||||
argparser.add_argument(
|
||||
'-p1', '--port1',
|
||||
metavar='P',
|
||||
default=2000,
|
||||
type=int,
|
||||
help='TCP port to listen to the first server (default: 2000)')
|
||||
argparser.add_argument(
|
||||
'--host2',
|
||||
metavar='H',
|
||||
default='127.0.0.1',
|
||||
help='IP of the second host server (default: 127.0.0.1)')
|
||||
argparser.add_argument(
|
||||
'-p2', '--port2',
|
||||
metavar='P',
|
||||
default=3000,
|
||||
type=int,
|
||||
help='TCP port to listen to the second server (default: 3000)')
|
||||
argparser.add_argument(
|
||||
'-i', '--images-to-disk',
|
||||
action='store_true',
|
||||
help='save images to disk')
|
||||
|
||||
args = argparser.parse_args()
|
||||
|
||||
logging_config = {
|
||||
'format': '%(levelname)s: %(message)s',
|
||||
'level': logging.DEBUG if args.debug else logging.INFO
|
||||
}
|
||||
if args.log:
|
||||
logging_config['filename'] = args.log
|
||||
logging_config['filemode'] = 'w+'
|
||||
logging.basicConfig(**logging_config)
|
||||
|
||||
logging.info('listening to 1st server at %s:%s', args.host1, args.port1)
|
||||
logging.info('listening to 2nd server at %s:%s', args.host2, args.port2)
|
||||
|
||||
while True:
|
||||
try:
|
||||
|
||||
run_carla_clients(args)
|
||||
|
||||
except TCPConnectionError as error:
|
||||
logging.error(error)
|
||||
time.sleep(1)
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
|
||||
try:
|
||||
main()
|
||||
except KeyboardInterrupt:
|
||||
print('\nCancelled by user. Bye!')
|
||||
|
|
@ -1,7 +1,7 @@
|
|||
#!/usr/bin/env python3
|
||||
|
||||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
# Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de
|
||||
# Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
# Barcelona (UAB).
|
||||
#
|
||||
# This work is licensed under the terms of the MIT license.
|
||||
# For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
46
README.md
46
README.md
|
|
@ -1,6 +1,10 @@
|
|||
CARLA Simulator
|
||||
===============
|
||||
|
||||
[](https://travis-ci.org/carla-simulator/carla)
|
||||
[](https://carla.readthedocs.io)
|
||||
[](https://waffle.io/carla-simulator/carla)
|
||||
|
||||
CARLA is an open-source simulator for autonomous driving research. CARLA has
|
||||
been developed from the ground up to support development, training, and
|
||||
validation of autonomous urban driving systems. In addition to open-source code
|
||||
|
|
@ -16,9 +20,15 @@ environmental conditions.
|
|||
For instructions on how to use and compile CARLA, check out
|
||||
[CARLA Documentation](http://carla.readthedocs.io).
|
||||
|
||||
If you want to benchmark your model in the same conditions as in our CoRL’17
|
||||
paper, check out
|
||||
[Benchmarking](http://carla.readthedocs.io/en/latest/benchmark/).
|
||||
|
||||
News
|
||||
----
|
||||
|
||||
- 25.01.2018 CARLA 0.7.1 released: [change log](https://github.com/carla-simulator/carla/blob/master/CHANGELOG.md#carla-071), [release](https://github.com/carla-simulator/carla/releases/tag/0.7.1).
|
||||
- 22.01.2018 Job opening: [C++ (UE4) Programmer](https://drive.google.com/open?id=1Hx0eUgpXl95d4IL9meEGhJECgSRos1T1).
|
||||
- 28.11.2017 CARLA 0.7.0 released: [change log](https://github.com/carla-simulator/carla/blob/master/CHANGELOG.md#carla-070), [release](https://github.com/carla-simulator/carla/releases/tag/0.7.0).
|
||||
- 15.11.2017 CARLA 0.6.0 released: [change log](https://github.com/carla-simulator/carla/blob/master/CHANGELOG.md#carla-060), [release](https://github.com/carla-simulator/carla/releases/tag/0.6.0).
|
||||
|
||||
|
|
@ -55,6 +65,42 @@ Felipe Codevilla, Antonio Lopez, Vladlen Koltun; PMLR 78:1-16
|
|||
}
|
||||
```
|
||||
|
||||
Building CARLA
|
||||
--------------
|
||||
|
||||
Use `git clone` or download the project from this page. Note that the master
|
||||
branch contains the latest fixes and features, for the latest stable code may be
|
||||
best to switch to the `stable` branch.
|
||||
|
||||
Then follow the instruction at [How to build on Linux][buildlink].
|
||||
|
||||
Unfortunately we don't have yet official instructions to build on other
|
||||
platforms, please check the progress for [Windows][issue21] and [Mac][issue150].
|
||||
|
||||
[buildlink]: http://carla.readthedocs.io/en/latest/how_to_build_on_linux
|
||||
[issue21]: https://github.com/carla-simulator/carla/issues/21
|
||||
[issue150]: https://github.com/carla-simulator/carla/issues/150
|
||||
|
||||
Contributing
|
||||
------------
|
||||
|
||||
Please take a look at our [Contribution guidelines][contriblink].
|
||||
|
||||
[contriblink]: http://carla.readthedocs.io/en/latest/CONTRIBUTING
|
||||
|
||||
F.A.Q.
|
||||
------
|
||||
|
||||
If you run into problems, check our
|
||||
[FAQ](http://carla.readthedocs.io/en/latest/faq/).
|
||||
|
||||
Jobs
|
||||
----
|
||||
|
||||
We are currently looking for a new programmer to join our team
|
||||
|
||||
* [C++ (UE4) Programmer](https://drive.google.com/open?id=1Hx0eUgpXl95d4IL9meEGhJECgSRos1T1)
|
||||
|
||||
License
|
||||
-------
|
||||
|
||||
|
|
|
|||
51
Rebuild.sh
51
Rebuild.sh
|
|
@ -1,6 +1,46 @@
|
|||
#!/bin/bash
|
||||
|
||||
################################################################################
|
||||
# Updates CARLA content.
|
||||
################################################################################
|
||||
|
||||
set -e
|
||||
|
||||
DOC_STRING="Update CARLA content to the latest version, to be run after 'git pull'."
|
||||
|
||||
USAGE_STRING="Usage: $0 [-h|--help] [--no-editor]"
|
||||
|
||||
# ==============================================================================
|
||||
# -- Parse arguments -----------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
LAUNCH_UE4_EDITOR=true
|
||||
|
||||
OPTS=`getopt -o h --long help,no-editor -n 'parse-options' -- "$@"`
|
||||
|
||||
if [ $? != 0 ] ; then echo "$USAGE_STRING" ; exit 2 ; fi
|
||||
|
||||
eval set -- "$OPTS"
|
||||
|
||||
while true; do
|
||||
case "$1" in
|
||||
--no-editor )
|
||||
LAUNCH_UE4_EDITOR=false;
|
||||
shift ;;
|
||||
-h | --help )
|
||||
echo "$DOC_STRING"
|
||||
echo "$USAGE_STRING"
|
||||
exit 1
|
||||
;;
|
||||
* )
|
||||
break ;;
|
||||
esac
|
||||
done
|
||||
|
||||
# ==============================================================================
|
||||
# -- Set up environment --------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
|
||||
pushd "$SCRIPT_DIR" >/dev/null
|
||||
|
||||
|
|
@ -60,8 +100,15 @@ set -e
|
|||
log "Build CarlaUE4 project..."
|
||||
make CarlaUE4Editor
|
||||
|
||||
log "Launching UE4Editor..."
|
||||
${UE4_ROOT}/Engine/Binaries/Linux/UE4Editor "${PWD}/CarlaUE4.uproject"
|
||||
if $LAUNCH_UE4_EDITOR ; then
|
||||
log "Launching UE4Editor..."
|
||||
${UE4_ROOT}/Engine/Binaries/Linux/UE4Editor "${PWD}/CarlaUE4.uproject"
|
||||
else
|
||||
echo ""
|
||||
echo "****************"
|
||||
echo "*** Success! ***"
|
||||
echo "****************"
|
||||
fi
|
||||
|
||||
popd >/dev/null
|
||||
|
||||
|
|
|
|||
24
Setup.sh
24
Setup.sh
|
|
@ -1,10 +1,14 @@
|
|||
#! /bin/bash
|
||||
|
||||
################################################################################
|
||||
# CARLA Util Setup
|
||||
# CARLA Setup.sh
|
||||
#
|
||||
# This downloads and compiles libc++. So we can build and compile our
|
||||
# dependencies with libc++ for linking against Unreal.
|
||||
# This script sets up the environment and dependencies for compiling CARLA on
|
||||
# Linux.
|
||||
#
|
||||
# 1) Download CARLA Content if necessary.
|
||||
# 2) Download and compile libc++.
|
||||
# 3) Download other third-party libraries and compile them with libc++.
|
||||
#
|
||||
# Thanks to the people at https://github.com/Microsoft/AirSim for providing the
|
||||
# important parts of this script.
|
||||
|
|
@ -12,6 +16,10 @@
|
|||
|
||||
set -e
|
||||
|
||||
# ==============================================================================
|
||||
# -- Set up environment --------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
|
||||
pushd "$SCRIPT_DIR" >/dev/null
|
||||
|
||||
|
|
@ -22,9 +30,6 @@ command -v clang++-3.9 >/dev/null 2>&1 || {
|
|||
exit 1;
|
||||
}
|
||||
|
||||
# Update content.
|
||||
./Update.sh
|
||||
|
||||
mkdir -p Util/Build
|
||||
pushd Util/Build >/dev/null
|
||||
|
||||
|
|
@ -161,6 +166,13 @@ fi
|
|||
|
||||
./Util/Protoc.sh
|
||||
|
||||
# ==============================================================================
|
||||
# -- Update CARLA Content ------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
||||
echo
|
||||
./Update.sh $@
|
||||
|
||||
# ==============================================================================
|
||||
# -- ...and we are done --------------------------------------------------------
|
||||
# ==============================================================================
|
||||
|
|
|
|||
|
|
@ -2,8 +2,8 @@
|
|||
ProjectID=675BF8694238308FA9368292CC440350
|
||||
ProjectName=CARLA UE4
|
||||
CompanyName=CVC
|
||||
CopyrightNotice="Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonomade Barcelona (UAB), and the INTEL Visual Computing Lab.This work is licensed under the terms of the MIT license.For a copy, see <https://opensource.org/licenses/MIT>."
|
||||
ProjectVersion=0.7.0
|
||||
CopyrightNotice="Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma de Barcelona (UAB). This work is licensed under the terms of the MIT license. For a copy, see <https://opensource.org/licenses/MIT>."
|
||||
ProjectVersion=0.7.1
|
||||
|
||||
[/Script/UnrealEd.ProjectPackagingSettings]
|
||||
BuildConfiguration=PPBC_Development
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
{
|
||||
"FileVersion": 3,
|
||||
"Version": 1,
|
||||
"VersionName": "0.7.0",
|
||||
"VersionName": "0.7.1",
|
||||
"FriendlyName": "CARLA",
|
||||
"Description": "",
|
||||
"Category": "Science",
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -104,7 +104,7 @@ void AVehicleSpawnerBase::SpawnVehicleAtSpawnPoint(
|
|||
Vehicle->SpawnDefaultController();
|
||||
auto Controller = GetController(Vehicle);
|
||||
if (Controller != nullptr) { // Sometimes fails...
|
||||
Controller->SetRandomEngine(GetRandomEngine());
|
||||
Controller->GetRandomEngine()->Seed(GetRandomEngine()->GenerateSeed());
|
||||
Controller->SetRoadMap(GetRoadMap());
|
||||
Controller->SetAutopilot(true);
|
||||
Vehicles.Add(Vehicle);
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -73,6 +73,8 @@ static void ClearQueue(std::queue<T> &Queue)
|
|||
AWheeledVehicleAIController::AWheeledVehicleAIController(const FObjectInitializer& ObjectInitializer) :
|
||||
Super(ObjectInitializer)
|
||||
{
|
||||
RandomEngine = CreateDefaultSubobject<URandomEngine>(TEXT("RandomEngine"));
|
||||
|
||||
PrimaryActorTick.bCanEverTick = true;
|
||||
PrimaryActorTick.TickGroup = TG_PrePhysics;
|
||||
}
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -98,14 +98,10 @@ public:
|
|||
/// @{
|
||||
public:
|
||||
|
||||
void SetRandomEngine(URandomEngine *InRandomEngine)
|
||||
{
|
||||
RandomEngine = InRandomEngine;
|
||||
}
|
||||
|
||||
UFUNCTION(Category = "Random Engine", BlueprintCallable)
|
||||
URandomEngine *GetRandomEngine()
|
||||
{
|
||||
check(RandomEngine != nullptr);
|
||||
return RandomEngine;
|
||||
}
|
||||
|
||||
|
|
@ -221,13 +217,13 @@ private:
|
|||
private:
|
||||
|
||||
UPROPERTY()
|
||||
ACarlaWheeledVehicle *Vehicle;
|
||||
ACarlaWheeledVehicle *Vehicle = nullptr;
|
||||
|
||||
UPROPERTY()
|
||||
URoadMap *RoadMap;
|
||||
URoadMap *RoadMap = nullptr;
|
||||
|
||||
UPROPERTY()
|
||||
URandomEngine *RandomEngine;
|
||||
URandomEngine *RandomEngine = nullptr;
|
||||
|
||||
UPROPERTY(VisibleAnywhere)
|
||||
bool bAutopilotEnabled = false;
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -0,0 +1,29 @@
|
|||
// Fill out your copyright notice in the Description page of Project Settings.
|
||||
|
||||
#pragma once
|
||||
|
||||
#include "CapturedLidarSegment.generated.h"
|
||||
|
||||
///
|
||||
/// Lidar segment captured by tick
|
||||
///
|
||||
USTRUCT()
|
||||
struct FCapturedLidarLaserSegment
|
||||
{
|
||||
GENERATED_USTRUCT_BODY()
|
||||
|
||||
UPROPERTY(VisibleAnywhere)
|
||||
TArray<FVector> Points;
|
||||
};
|
||||
|
||||
USTRUCT()
|
||||
struct FCapturedLidarSegment
|
||||
{
|
||||
GENERATED_USTRUCT_BODY()
|
||||
|
||||
UPROPERTY(VisibleAnywhere)
|
||||
float HorizontalAngle = 0;
|
||||
|
||||
UPROPERTY(VisibleAnywhere)
|
||||
TArray<FCapturedLidarLaserSegment> LidarLasersSegments;
|
||||
};
|
||||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
@ -176,7 +176,8 @@ void ACarlaGameModeBase::BeginPlay()
|
|||
VehicleSpawner->SetSeed(CarlaSettings.SeedVehicles);
|
||||
VehicleSpawner->SetRoadMap(RoadMap);
|
||||
if (PlayerController != nullptr) {
|
||||
PlayerController->SetRandomEngine(VehicleSpawner->GetRandomEngine());
|
||||
PlayerController->GetRandomEngine()->Seed(
|
||||
VehicleSpawner->GetRandomEngine()->GenerateSeed());
|
||||
}
|
||||
} else {
|
||||
UE_LOG(LogCarla, Error, TEXT("Missing vehicle spawner actor!"));
|
||||
|
|
@ -224,6 +225,10 @@ void ACarlaGameModeBase::AttachCaptureCamerasToPlayer()
|
|||
for (const auto &Item : Settings.CameraDescriptions) {
|
||||
PlayerController->AddSceneCaptureCamera(Item.Value, OverridePostProcessParameters);
|
||||
}
|
||||
|
||||
for (const auto &Item : Settings.LidarDescriptions) {
|
||||
PlayerController->AddSceneCaptureLidar(Item.Value);
|
||||
}
|
||||
}
|
||||
|
||||
void ACarlaGameModeBase::TagActorsForSemanticSegmentation()
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
// Copyright (c) 2017 Computer Vision Center (CVC) at the Universitat Autonoma
|
||||
// de Barcelona (UAB), and the INTEL Visual Computing Lab.
|
||||
// de Barcelona (UAB).
|
||||
//
|
||||
// This work is licensed under the terms of the MIT license.
|
||||
// For a copy, see <https://opensource.org/licenses/MIT>.
|
||||
|
|
|
|||
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in New Issue