p85017694
/
software_program
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

修改ReadMeFirst，同时删除opensrc中没有用到的代码

This commit is contained in:
Chanlen 2018-06-13 00:36:18 +08:00
parent 0fbf5b0c9f
commit 07ad4c0b33
11 changed files with 22 additions and 2219 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
ReadMeFirst.txt
View File

@ -4,18 +4,26 @@ This is a brief introduce of our software program.
|---->幼儿实时监视预警系统-汇报PPT:课程实践过程中产生的PPT文件
|---->幼儿实时监视预警系统-项目简介.pptx: 项目初期的一个简单介绍的PPT文件
|---->幼儿实时监视预警系统 -第一次迭代汇报.pptx:第一学期结课时汇报使用的PPT文件
|---->幼儿实时监视预警系统-项目进度报告.pptx
|---->幼儿实时监视预警系统-项目进度报告.pptx: 第二学期开学时进度汇报的PPT文件
|---->幼儿实时监视预警系统-课程实践汇报.pptx:课程实践基本完成使用的汇报PPT文件
|---->幼儿实时监视预警系统宣传材料
|---->幼儿实时监视预警系统-实践自评表.xlsx
|---->幼儿实时监视预警系统-软件需求构想与描述.docx
|---->幼儿实时监视预警系统-软件需求规格说明书.doc
|---->幼儿实时监视预警系统.png: 项目的宣传彩页
|---->幼儿实时监视预警系统-实践自评表.xlsx: 项目的实践自评表
|---->幼儿实时监视预警系统-软件需求构想与描述.docx: 软件需求与构想文档
|---->幼儿实时监视预警系统-软件需求规格说明书.doc: 软件的需求规格说明书
model: 软件设计中产生的模型
|---->幼儿实时监视预警系统 - 类设计模型.docx
|---->幼儿实时监视预警系统- 数据库设计模型.docx
|---->幼儿实时监视预警系统- 用例设计模型.vsdx
|---->幼儿实时监视预警系统- 界面类图.pdf
|---->幼儿实时监视预警系统- 类设计模型.vsdx
|---->幼儿实时监视预警系统- 软件体系结构图.docx
src: 存储软件设计过程中的相关代码
|---->dangerDetect: 通过图像识别的方式检测幼儿是否可能受到危险
|---->EBaby: 手机端的代码
|---->Socket: 手机和上网本通信部分代码
openSrc：使用的开源软件代码
draft: 文档撰写过程中的原材料，不一一枚举
|---->cryDetect: 哭泣检测部分相关代码
|---->openSrc：使用的开源软件代码
|---->ReadMeFirst.txt: 对于代码的说明

7
src/ReadMeFirst.txt
View File

@ -1 +1,6 @@
当前代码为软件原型的实现代码，还没有做精化
dangerDetect: 通过图像识别的方式检测幼儿是否可能受到危险
EBaby: 手机端的代码
Socket: 手机和上网本通信部分代码
cryDetect: 哭泣检测部分相关代码
openSrc：使用的开源软件代码
说明：openSrc为项目引入的开源代码，其余部分为自己编写的代码

272
src/openSrc/xfei_asr/src/asr_sample.c~
View File

@ -1,272 +0,0 @@
/*
* 语音识别（Automatic Speech Recognition）技术能够从语音中识别出特定的命令词或语句模式。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#define BUFFER_SIZE 2048
#define HINTS_SIZE 100
#define GRAMID_LEN 128
#define FRAME_LEN 640
int get_grammar_id(char* grammar_id, unsigned int id_len)
{
FILE* fp = NULL;
char* grammar = NULL;
unsigned int grammar_len = 0;
unsigned int read_len = 0;
const char* ret_id = NULL;
unsigned int ret_id_len = 0;
int ret = -1;
if (NULL == grammar_id)
goto grammar_exit;
fp = fopen("gm_continuous_digit.abnf", "rb");
if (NULL == fp)
{
printf("\nopen grammar file failed!\n");
goto grammar_exit;
}
fseek(fp, 0, SEEK_END);
grammar_len = ftell(fp); //获取语法文件大小
fseek(fp, 0, SEEK_SET);
grammar = (char*)malloc(grammar_len + 1);
if (NULL == grammar)
{
printf("\nout of memory!\n");
goto grammar_exit;
}
read_len = fread((void *)grammar, 1, grammar_len, fp); //读取语法内容
if (read_len != grammar_len)
{
printf("\nread grammar error!\n");
goto grammar_exit;
}
grammar[grammar_len] = '\0';
ret_id = MSPUploadData("usergram", grammar, grammar_len, "dtt = abnf, sub = asr", &ret); //上传语法
if (MSP_SUCCESS != ret)
{
printf("\nMSPUploadData failed, error code: %d.\n", ret);
goto grammar_exit;
}
ret_id_len = strlen(ret_id);
if (ret_id_len >= id_len)
{
printf("\nno enough buffer for grammar_id!\n");
goto grammar_exit;
}
strncpy(grammar_id, ret_id, ret_id_len);
printf("grammar_id: \"%s\" \n", grammar_id); //下次可以直接使用该ID，不必重复上传语法。
grammar_exit:
if (NULL != fp)
{
fclose(fp);
fp = NULL;
}
if (NULL!= grammar)
{
free(grammar);
grammar = NULL;
}
return ret;
}
void run_asr(const char* audio_file, const char* params, char* grammar_id)
{
const char* session_id = NULL;
char rec_result[BUFFER_SIZE] = {'\0'};
char hints[HINTS_SIZE] = {'\0'}; //hints为结束本次会话的原因描述，由用户自定义
unsigned int total_len = 0;
int aud_stat = MSP_AUDIO_SAMPLE_CONTINUE; //音频状态
int ep_stat = MSP_EP_LOOKING_FOR_SPEECH; //端点检测
int rec_stat = MSP_REC_STATUS_SUCCESS; //识别状态
int errcode = MSP_SUCCESS;
FILE* f_pcm = NULL;
char* p_pcm = NULL;
long pcm_count = 0;
long pcm_size = 0;
long read_size = 0;
if (NULL == audio_file)
goto asr_exit;
f_pcm = fopen(audio_file, "rb");
if (NULL == f_pcm)
{
printf("\nopen [%s] failed!\n", audio_file);
goto asr_exit;
}
fseek(f_pcm, 0, SEEK_END);
pcm_size = ftell(f_pcm); //获取音频文件大小
fseek(f_pcm, 0, SEEK_SET);
p_pcm = (char*)malloc(pcm_size);
if (NULL == p_pcm)
{
printf("\nout of memory!\n");
goto asr_exit;
}
read_size = fread((void *)p_pcm, 1, pcm_size, f_pcm); //读取音频文件内容
if (read_size != pcm_size)
{
printf("\nread [%s] failed!\n", audio_file);
goto asr_exit;
}
printf("\n开始语音识别 ...\n");
session_id = QISRSessionBegin(grammar_id, params, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRSessionBegin failed, error code:%d\n", errcode);
goto asr_exit;
}
while (1)
{
unsigned int len = 10 * FRAME_LEN; // 每次写入200ms音频(16k，16bit)：1帧音频20ms，10帧=200ms。16k采样率的16位音频，一帧的大小为640Byte
int ret = 0;
if (pcm_size < 2 * len)
len = pcm_size;
if (len <= 0)
break;
aud_stat = MSP_AUDIO_SAMPLE_CONTINUE;
if (0 == pcm_count)
aud_stat = MSP_AUDIO_SAMPLE_FIRST;
printf(">");
ret = QISRAudioWrite(session_id, (const void *)&p_pcm[pcm_count], len, aud_stat, &ep_stat, &rec_stat);
if (MSP_SUCCESS != ret)
{
printf("\nQISRAudioWrite failed, error code:%d\n",ret);
goto asr_exit;
}
pcm_count += (long)len;
pcm_size -= (long)len;
if (MSP_EP_AFTER_SPEECH == ep_stat)
break;
usleep(200*1000); //模拟人说话时间间隙，10帧的音频长度为200ms
}
errcode = QISRAudioWrite(session_id, NULL, 0, MSP_AUDIO_SAMPLE_LAST, &ep_stat, &rec_stat);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRAudioWrite failed, error code:%d\n",errcode);
goto asr_exit;
}
while (MSP_REC_STATUS_COMPLETE != rec_stat)
{
const char *rslt = QISRGetResult(session_id, &rec_stat, 0, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRGetResult failed, error code: %d\n", errcode);
goto asr_exit;
}
if (NULL != rslt)
{
unsigned int rslt_len = strlen(rslt);
total_len += rslt_len;
if (total_len >= BUFFER_SIZE)
{
printf("\nno enough buffer for rec_result !\n");
goto asr_exit;
}
strncat(rec_result, rslt, rslt_len);
}
usleep(150*1000); //防止频繁占用CPU
}
printf("\n语音识别结束\n");
printf("=============================================================\n");
printf("%s",rec_result);
printf("=============================================================\n");
asr_exit:
if (NULL != f_pcm)
{
fclose(f_pcm);
f_pcm = NULL;
}
if (NULL != p_pcm)
{
free(p_pcm);
p_pcm = NULL;
}
QISRSessionEnd(session_id, hints);
}
int main(int argc, char* argv[])
{
int ret = MSP_SUCCESS;
const char* login_params = "appid = 5afda937, work_dir = ."; //登录参数,appid与msc库绑定,请勿随意改动
/*
* sub: 请求业务类型
* result_type: 识别结果格式
* result_encoding: 结果编码格式
*
* 详细参数说明请参阅《讯飞语音云MSC--API文档》
*/
const char* session_begin_params = "sub = asr, result_type = plain, result_encoding = utf8";
char* grammar_id = NULL;
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params); //第一个参数是用户名，第二个参数是密码，均传NULL即可，第三个参数是登录参数
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed, error code: %d.\n",ret);
goto exit; //登录失败，退出登录
}
printf("\n##################################################\n");
printf("## 语音识别（Automatic Speech Recognition）技术 ##\n");
printf("## 能够从语音中识别出特定的命令词或语句模式。 ##\n");
printf("##################################################\n\n");
grammar_id = (char*)malloc(GRAMID_LEN);
if (NULL == grammar_id)
{
printf("out of memory !\n");
goto exit;
}
memset(grammar_id, 0, GRAMID_LEN);
printf("上传语法 ...\n");
ret = get_grammar_id(grammar_id, GRAMID_LEN);
if (MSP_SUCCESS != ret)
goto exit;
printf("上传语法成功\n");
run_asr("wav/iflytek01.wav", session_begin_params, grammar_id); //iflytek01对应的音频内容：“18012345678”
exit:
if (NULL != grammar_id)
{
free(grammar_id);
grammar_id = NULL;
}
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}

344
src/openSrc/xfei_asr/src/cryDetect.c~
View File

@ -1,344 +0,0 @@
/*
* 语音听写(iFly Auto Transform)技术能够实时地将语音转换成对应的文字。
*/
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#include "speech_recognizer.h"
#include <iconv.h>
#define FRAME_LEN 640
#define BUFFER_SIZE 4096
/* Upload User words */
static int upload_userwords()
{
char* userwords = NULL;
size_t len = 0;
size_t read_len = 0;
FILE* fp = NULL;
int ret = -1;
fp = fopen("userwords.txt", "rb");
if (NULL == fp)
{
printf("\nopen [userwords.txt] failed! \n");
goto upload_exit;
}
fseek(fp, 0, SEEK_END);
len = ftell(fp);
fseek(fp, 0, SEEK_SET);
userwords = (char*)malloc(len + 1);
if (NULL == userwords)
{
printf("\nout of memory! \n");
goto upload_exit;
}
read_len = fread((void*)userwords, 1, len, fp);
if (read_len != len)
{
printf("\nread [userwords.txt] failed!\n");
goto upload_exit;
}
userwords[len] = '\0';
MSPUploadData("userwords", userwords, len, "sub = uup, dtt = userword", &ret); //ÉÏ´«Óû§´Ê±í
if (MSP_SUCCESS != ret)
{
printf("\nMSPUploadData failed ! errorCode: %d \n", ret);
goto upload_exit;
}
upload_exit:
if (NULL != fp)
{
fclose(fp);
fp = NULL;
}
if (NULL != userwords)
{
free(userwords);
userwords = NULL;
}
return ret;
}
void alert()
{
//gnome-terminal -x bash -c "rostopic pub /chatter std_msgs/String "alert" ; exec bash\"
system("gnome-terminal -x bash -c \"rostopic pub /chatter std_msgs/String \"alert\" ; exec bash\"");
system("play ~/alert.mp3");
}
static void show_result(char *string, char is_over)
{
printf("\rResult:[ %s ]", string);
int strIndex = 0; //字符串指针的位置
int countTimes = 0;//记录出现关键字的次数
int len = strlen(string);
for(strIndex = 0; strIndex < len -1;strIndex++)
{
if((string[strIndex] == 'h'&& string[strIndex+1] == 'a') || (string[strIndex] == 'w'&& string[strIndex+1] == 'a'))
{
countTimes ++;
if(countTimes == 5)
{
alert();
break;
}
}
}
if(is_over)
putchar('\n');
}
static char *g_result = NULL;
static unsigned int g_buffersize = BUFFER_SIZE;
void on_result(const char *result, char is_last)
{
if (result) {
size_t left = g_buffersize - 1 - strlen(g_result);
size_t size = strlen(result);
if (left < size) {
g_result = (char*)realloc(g_result, g_buffersize + BUFFER_SIZE);
if (g_result)
g_buffersize += BUFFER_SIZE;
else {
printf("mem alloc failed\n");
return;
}
}
strncat(g_result, result, size);
show_result(g_result, is_last);
}
}
void on_speech_begin()
{
if (g_result)
{
free(g_result);
}
g_result = (char*)malloc(BUFFER_SIZE);
g_buffersize = BUFFER_SIZE;
memset(g_result, 0, g_buffersize);
printf("Start Listening...\n");
}
void on_speech_end(int reason)
{
if (reason == END_REASON_VAD_DETECT)
printf("\nSpeaking done \n");
else
printf("\nRecognizer error %d\n", reason);
}
/* demo send audio data from a file */
static void demo_file(const char* audio_file, const char* session_begin_params)
{
int errcode = 0;
FILE* f_pcm = NULL;
char* p_pcm = NULL;
unsigned long pcm_count = 0;
unsigned long pcm_size = 0;
unsigned long read_size = 0;
struct speech_rec iat;
struct speech_rec_notifier recnotifier = {
on_result,
on_speech_begin,
on_speech_end
};
if (NULL == audio_file)
goto iat_exit;
f_pcm = fopen(audio_file, "rb");
if (NULL == f_pcm)
{
printf("\nopen [%s] failed! \n", audio_file);
goto iat_exit;
}
fseek(f_pcm, 0, SEEK_END);
pcm_size = ftell(f_pcm);
fseek(f_pcm, 0, SEEK_SET);
p_pcm = (char *)malloc(pcm_size);
if (NULL == p_pcm)
{
printf("\nout of memory! \n");
goto iat_exit;
}
read_size = fread((void *)p_pcm, 1, pcm_size, f_pcm);
if (read_size != pcm_size)
{
printf("\nread [%s] error!\n", audio_file);
goto iat_exit;
}
errcode = sr_init(&iat, session_begin_params, SR_USER, &recnotifier);
if (errcode) {
printf("speech recognizer init failed : %d\n", errcode);
goto iat_exit;
}
errcode = sr_start_listening(&iat);
if (errcode) {
printf("\nsr_start_listening failed! error code:%d\n", errcode);
goto iat_exit;
}
while (1)
{
unsigned int len = 10 * FRAME_LEN; /* 200ms audio */
int ret = 0;
if (pcm_size < 2 * len)
len = pcm_size;
if (len <= 0)
break;
ret = sr_write_audio_data(&iat, &p_pcm[pcm_count], len);
if (0 != ret)
{
printf("\nwrite audio data failed! error code:%d\n", ret);
goto iat_exit;
}
pcm_count += (long)len;
pcm_size -= (long)len;
}
errcode = sr_stop_listening(&iat);
if (errcode) {
printf("\nsr_stop_listening failed! error code:%d \n", errcode);
goto iat_exit;
}
iat_exit:
if (NULL != f_pcm)
{
fclose(f_pcm);
f_pcm = NULL;
}
if (NULL != p_pcm)
{
free(p_pcm);
p_pcm = NULL;
}
sr_stop_listening(&iat);
sr_uninit(&iat);
}
/* demo recognize the audio from microphone */
static void demo_mic(const char* session_begin_params)
{
int errcode;
int i = 0;
struct speech_rec iat;
struct speech_rec_notifier recnotifier = {
on_result,
on_speech_begin,
on_speech_end
};
errcode = sr_init(&iat, session_begin_params, SR_MIC, &recnotifier);
if (errcode) {
printf("speech recognizer init failed\n");
return;
}
errcode = sr_start_listening(&iat);
if (errcode) {
printf("start listen failed %d\n", errcode);
}
/* demo 15 seconds recording */
while(i++ < 15)
sleep(1);
errcode = sr_stop_listening(&iat);
if (errcode) {
printf("stop listening failed %d\n", errcode);
}
sr_uninit(&iat);
}
/* main thread: start/stop record ; query the result of recgonization.
* record thread: record callback(data write)
* helper thread: ui(keystroke detection)
*/
int main(int argc, char* argv[])
{
//FILE *fp = fopen("result.txt","w");
int ret = MSP_SUCCESS;
int upload_on = 1; /* whether upload the user word */
/* login params, please do keep the appid correct */
const char* login_params = "appid = 5afd8086, work_dir = .";
int aud_src = 0; /* from mic or file */
/*
* See "iFlytek MSC Reference Manual" "sub = iat, domain = iat, language = zh_cn, "
*/
const char* session_begin_params =
"sub = iat, domain = iat, language = en_us, "
"accent = mandarin, sample_rate = 16000, "
"result_type = plain, result_encoding = utf8";
/* Login first. the 1st arg is username, the 2nd arg is password
* just set them as NULL. the 3rd arg is login paramertes
* */
ret = MSPLogin(NULL, NULL, login_params);
if (MSP_SUCCESS != ret) {
printf("MSPLogin failed , Error code %d.\n",ret);
goto exit; // login fail, exit the program
}
printf("Want to upload the user words ? \n0: No.\n1: Yes\n");
scanf("%d", &upload_on);
if (upload_on)
{
printf("Uploading the user words ...\n");
ret = upload_userwords();
if (MSP_SUCCESS != ret)
goto exit;
printf("Uploaded successfully\n");
}
printf("Where the audio comes from?\n"
"0: From a audio file.\n1: From microphone.\n");
scanf("%d", &aud_src);
if(aud_src != 0) {
printf("Demo recognizing the speech from microphone\n");
printf("Speak in 15 seconds\n");
demo_mic(session_begin_params);
printf("15 sec passed\n");
}
else {
//printf("Please type the route of the source audio:\n");
printf("Demo recgonizing the speech from a recorded audio file\n");
demo_file("wav/iflytek02.wav", session_begin_params);
}
exit:
MSPLogout(); // Logout...
//fclose(fp);
return 0;
}

253
src/openSrc/xfei_asr/src/iat_publish_speak.cpp~
View File

@ -1,253 +0,0 @@
/*
* 语音听写(iFly Auto Transform)技术能够实时地将语音转换成对应的文字。
*/
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#include "speech_recognizer.h"
#include <iconv.h>
#include "ros/ros.h"
#include "std_msgs/String.h"
#define FRAME_LEN 640
#define BUFFER_SIZE 4096
int flag = 0 ;
int flag_ok = 0 ;
int flag_no = 0 ;
/* Upload User words */
static int upload_userwords()
{
char* userwords = NULL;
size_t len = 0;
size_t read_len = 0;
FILE* fp = NULL;
int ret = -1;
fp = fopen("userwords.txt", "rb");
if (NULL == fp)
{
printf("\nopen [userwords.txt] failed! \n");
goto upload_exit;
}
fseek(fp, 0, SEEK_END);
len = ftell(fp);
fseek(fp, 0, SEEK_SET);
userwords = (char*)malloc(len + 1);
if (NULL == userwords)
{
printf("\nout of memory! \n");
goto upload_exit;
}
read_len = fread((void*)userwords, 1, len, fp);
if (read_len != len)
{
printf("\nread [userwords.txt] failed!\n");
goto upload_exit;
}
userwords[len] = '\0';
MSPUploadData("userwords", userwords, len, "sub = uup, dtt = userword", &ret); //ÉÏ´«Óû§´Ê±í
if (MSP_SUCCESS != ret)
{
printf("\nMSPUploadData failed ! errorCode: %d \n", ret);
goto upload_exit;
}
upload_exit:
if (NULL != fp)
{
fclose(fp);
fp = NULL;
}
if (NULL != userwords)
{
free(userwords);
userwords = NULL;
}
return ret;
}
static void show_result(char *string, char is_over)
{
flag_ok=1;
printf("\rResult:[ %s ]", string);
if(is_over)
putchar('\n');
}
static char *g_result = NULL;
static unsigned int g_buffersize = BUFFER_SIZE;
void on_result(const char *result, char is_last)
{
if (result) {
size_t left = g_buffersize - 1 - strlen(g_result);
size_t size = strlen(result);
if (left < size) {
g_result = (char*)realloc(g_result, g_buffersize + BUFFER_SIZE);
if (g_result)
g_buffersize += BUFFER_SIZE;
else {
printf("mem alloc failed\n");
return;
}
}
strncat(g_result, result, size);
show_result(g_result, is_last);
}
}
void on_speech_begin()
{
if (g_result)
{
free(g_result);
}
g_result = (char*)malloc(BUFFER_SIZE);
g_buffersize = BUFFER_SIZE;
memset(g_result, 0, g_buffersize);
printf("Start Listening...\n");
}
void on_speech_end(int reason)
{
if (reason == END_REASON_VAD_DETECT)
printf("\nSpeaking done \n");
else
printf("\nRecognizer error %d\n", reason);
}
/* demo recognize the audio from microphone */
static void demo_mic(const char* session_begin_params)
{
int errcode;
int i = 0;
struct speech_rec iat;
struct speech_rec_notifier recnotifier = {
on_result,
on_speech_begin,
on_speech_end
};
errcode = sr_init(&iat, session_begin_params, SR_MIC, &recnotifier);
if (errcode) {
flag_no=1;
printf("speech recognizer init failed\n");
return;
}
errcode = sr_start_listening(&iat);
if (errcode) {
printf("start listen failed %d\n", errcode);
}
/* demo 15 seconds recording */
while(i++ < 15)
sleep(1);
errcode = sr_stop_listening(&iat);
if (errcode) {
flag_no=1;
printf("stop listening failed %d\n", errcode);
}
sr_uninit(&iat);
}
/* main thread: start/stop record ; query the result of recgonization.
* record thread: record callback(data write)
* helper thread: ui(keystroke detection)
*/
void WakeUp(const std_msgs::String::ConstPtr& msg)
{
printf("waking up\r\n");
// printf("%s", *msg->data.c_str());
usleep(700*1000);
flag=1;
}
int main(int argc, char* argv[])
{
ros::init(argc, argv, "xfspeech");
ros::NodeHandle n;
ros::Publisher pub = n.advertise<std_msgs::String>("xfspeech", 1000);
ros::Rate loop_rate(10);
ros::Subscriber sbu = n.subscribe("xfwakeup", 1000, WakeUp);
ros::Publisher pub1 = n.advertise<std_msgs::String>("xfwords", 1000);
ros::Publisher pub2 = n.advertise<std_msgs::String>("xfspeech", 1000);
int count=0;
while(ros::ok())
{
if (flag){
int ret = MSP_SUCCESS;
const char* login_params = "appid = 5afda937, work_dir = .";
const char* session_begin_params =
"sub = iat, domain = iat, language = zh_cn, "
"accent = mandarin, sample_rate = 16000, "
"result_type = plain, result_encoding = utf8";
ret = MSPLogin(NULL, NULL, login_params);
if(MSP_SUCCESS != ret){
MSPLogout();
printf("MSPLogin failed , Error code %d.\n",ret);
}
printf("Demo recognizing the speech from microphone\n");
printf("Speak in 15 seconds\n");
demo_mic(session_begin_params);
printf("15 sec passed\n");
flag=0;
MSPLogout();
}
if(flag_ok){
flag_ok=0;
std_msgs::String msg;
msg.data = g_result;
pub2.publish(msg);
}
if(flag_no){
flag_no=0;
std_msgs::String msg;
msg.data = "Sorry Please do it again";
pub1.publish(msg);
}
ros::spinOnce();
loop_rate.sleep();
count++;
printf("c:%d \r\n", count);
}
exit:
MSPLogout(); // Logout...
return 0;
}

341
src/openSrc/xfei_asr/src/iat_record.c~
View File

@ -1,341 +0,0 @@
/*
* 语音听写(iFly Auto Transform)技术能够实时地将语音转换成对应的文字。
*/
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#include "speech_recognizer.h"
#include <iconv.h>
#define FRAME_LEN 640
#define BUFFER_SIZE 4096
/* Upload User words */
static int upload_userwords()
{
char* userwords = NULL;
size_t len = 0;
size_t read_len = 0;
FILE* fp = NULL;
int ret = -1;
fp = fopen("userwords.txt", "rb");
if (NULL == fp)
{
printf("\nopen [userwords.txt] failed! \n");
goto upload_exit;
}
fseek(fp, 0, SEEK_END);
len = ftell(fp);
fseek(fp, 0, SEEK_SET);
userwords = (char*)malloc(len + 1);
if (NULL == userwords)
{
printf("\nout of memory! \n");
goto upload_exit;
}
read_len = fread((void*)userwords, 1, len, fp);
if (read_len != len)
{
printf("\nread [userwords.txt] failed!\n");
goto upload_exit;
}
userwords[len] = '\0';
MSPUploadData("userwords", userwords, len, "sub = uup, dtt = userword", &ret); //ÉÏ´«Óû§´Ê±í
if (MSP_SUCCESS != ret)
{
printf("\nMSPUploadData failed ! errorCode: %d \n", ret);
goto upload_exit;
}
upload_exit:
if (NULL != fp)
{
fclose(fp);
fp = NULL;
}
if (NULL != userwords)
{
free(userwords);
userwords = NULL;
}
return ret;
}
static void show_result(char *string, char is_over)
{
char buf[100] = {0};
FILE* fd = fopen("result.txt", "w");
printf("\rResult:[ %s ]", string);
fprintf(fd,"%s\n", string);
int strPos = 0;
int countTimes = 0;
for(strPos = 0; strPos < strlen(string) -1;strPos++)
{
if((string[strPos] == 'h'&& string[strPos+1] == 'a')
|| (string[strPos] == 'w'&& string[strPos+1] == 'a'))
countTimes ++;
if(countTimes == 4)
{
//system("play ~/alert.mp3");
system("rostopic pub /chatter std_msgs/String \"alert\" ");
break;
}
}
if(is_over)
putchar('\n');
fclose(fd);
}
static char *g_result = NULL;
static unsigned int g_buffersize = BUFFER_SIZE;
void on_result(const char *result, char is_last)
{
if (result) {
size_t left = g_buffersize - 1 - strlen(g_result);
size_t size = strlen(result);
if (left < size) {
g_result = (char*)realloc(g_result, g_buffersize + BUFFER_SIZE);
if (g_result)
g_buffersize += BUFFER_SIZE;
else {
printf("mem alloc failed\n");
return;
}
}
strncat(g_result, result, size);
show_result(g_result, is_last);
}
}
void on_speech_begin()
{
if (g_result)
{
free(g_result);
}
g_result = (char*)malloc(BUFFER_SIZE);
g_buffersize = BUFFER_SIZE;
memset(g_result, 0, g_buffersize);
printf("Start Listening...\n");
}
void on_speech_end(int reason)
{
if (reason == END_REASON_VAD_DETECT)
printf("\nSpeaking done \n");
else
printf("\nRecognizer error %d\n", reason);
}
/* demo send audio data from a file */
static void demo_file(const char* audio_file, const char* session_begin_params)
{
int errcode = 0;
FILE* f_pcm = NULL;
char* p_pcm = NULL;
unsigned long pcm_count = 0;
unsigned long pcm_size = 0;
unsigned long read_size = 0;
struct speech_rec iat;
struct speech_rec_notifier recnotifier = {
on_result,
on_speech_begin,
on_speech_end
};
if (NULL == audio_file)
goto iat_exit;
f_pcm = fopen(audio_file, "rb");
if (NULL == f_pcm)
{
printf("\nopen [%s] failed! \n", audio_file);
goto iat_exit;
}
fseek(f_pcm, 0, SEEK_END);
pcm_size = ftell(f_pcm);
fseek(f_pcm, 0, SEEK_SET);
p_pcm = (char *)malloc(pcm_size);
if (NULL == p_pcm)
{
printf("\nout of memory! \n");
goto iat_exit;
}
read_size = fread((void *)p_pcm, 1, pcm_size, f_pcm);
if (read_size != pcm_size)
{
printf("\nread [%s] error!\n", audio_file);
goto iat_exit;
}
errcode = sr_init(&iat, session_begin_params, SR_USER, &recnotifier);
if (errcode) {
printf("speech recognizer init failed : %d\n", errcode);
goto iat_exit;
}
errcode = sr_start_listening(&iat);
if (errcode) {
printf("\nsr_start_listening failed! error code:%d\n", errcode);
goto iat_exit;
}
while (1)
{
unsigned int len = 10 * FRAME_LEN; /* 200ms audio */
int ret = 0;
if (pcm_size < 2 * len)
len = pcm_size;
if (len <= 0)
break;
ret = sr_write_audio_data(&iat, &p_pcm[pcm_count], len);
if (0 != ret)
{
printf("\nwrite audio data failed! error code:%d\n", ret);
goto iat_exit;
}
pcm_count += (long)len;
pcm_size -= (long)len;
}
errcode = sr_stop_listening(&iat);
if (errcode) {
printf("\nsr_stop_listening failed! error code:%d \n", errcode);
goto iat_exit;
}
iat_exit:
if (NULL != f_pcm)
{
fclose(f_pcm);
f_pcm = NULL;
}
if (NULL != p_pcm)
{
free(p_pcm);
p_pcm = NULL;
}
sr_stop_listening(&iat);
sr_uninit(&iat);
}
/* demo recognize the audio from microphone */
static void demo_mic(const char* session_begin_params)
{
int errcode;
int i = 0;
struct speech_rec iat;
struct speech_rec_notifier recnotifier = {
on_result,
on_speech_begin,
on_speech_end
};
errcode = sr_init(&iat, session_begin_params, SR_MIC, &recnotifier);
if (errcode) {
printf("speech recognizer init failed\n");
return;
}
errcode = sr_start_listening(&iat);
if (errcode) {
printf("start listen failed %d\n", errcode);
}
/* demo 15 seconds recording */
while(i++ < 15)
sleep(1);
errcode = sr_stop_listening(&iat);
if (errcode) {
printf("stop listening failed %d\n", errcode);
}
sr_uninit(&iat);
}
/* main thread: start/stop record ; query the result of recgonization.
* record thread: record callback(data write)
* helper thread: ui(keystroke detection)
*/
int main(int argc, char* argv[])
{
//FILE *fp = fopen("result.txt","w");
int ret = MSP_SUCCESS;
int upload_on = 1; /* whether upload the user word */
/* login params, please do keep the appid correct */
const char* login_params = "appid = 5afd8086, work_dir = .";
int aud_src = 0; /* from mic or file */
/*
* See "iFlytek MSC Reference Manual" "sub = iat, domain = iat, language = zh_cn, "
*/
const char* session_begin_params =
"sub = iat, domain = iat, language = en_us, "
"accent = mandarin, sample_rate = 16000, "
"result_type = plain, result_encoding = utf8";
/* Login first. the 1st arg is username, the 2nd arg is password
* just set them as NULL. the 3rd arg is login paramertes
* */
ret = MSPLogin(NULL, NULL, login_params);
if (MSP_SUCCESS != ret) {
printf("MSPLogin failed , Error code %d.\n",ret);
goto exit; // login fail, exit the program
}
printf("Want to upload the user words ? \n0: No.\n1: Yes\n");
scanf("%d", &upload_on);
if (upload_on)
{
printf("Uploading the user words ...\n");
ret = upload_userwords();
if (MSP_SUCCESS != ret)
goto exit;
printf("Uploaded successfully\n");
}
printf("Where the audio comes from?\n"
"0: From a audio file.\n1: From microphone.\n");
scanf("%d", &aud_src);
if(aud_src != 0) {
printf("Demo recognizing the speech from microphone\n");
printf("Speak in 15 seconds\n");
demo_mic(session_begin_params);
printf("15 sec passed\n");
}
else {
//printf("Please type the route of the source audio:\n");
printf("Demo recgonizing the speech from a recorded audio file\n");
demo_file("wav/iflytek02.wav", session_begin_params);
}
exit:
MSPLogout(); // Logout...
//fclose(fp);
return 0;
}

275
src/openSrc/xfei_asr/src/iat_sample.c~
View File

@ -1,275 +0,0 @@
/*
* 语音听写(iFly Auto Transform)技术能够实时地将语音转换成对应的文字。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#define BUFFER_SIZE 4096
#define FRAME_LEN 640
#define HINTS_SIZE 100
/* 上传用户词表 */
int upload_userwords()
{
char* userwords = NULL;
unsigned int len = 0;
unsigned int read_len = 0;
FILE* fp = NULL;
int ret = -1;
fp = fopen("userwords.txt", "rb");
if (NULL == fp)
{
printf("\nopen [userwords.txt] failed! \n");
goto upload_exit;
}
fseek(fp, 0, SEEK_END);
len = ftell(fp); //获取音频文件大小
fseek(fp, 0, SEEK_SET);
userwords = (char*)malloc(len + 1);
if (NULL == userwords)
{
printf("\nout of memory! \n");
goto upload_exit;
}
read_len = fread((void*)userwords, 1, len, fp); //读取用户词表内容
if (read_len != len)
{
printf("\nread [userwords.txt] failed!\n");
goto upload_exit;
}
userwords[len] = '\0';
MSPUploadData("userwords", userwords, len, "sub = uup, dtt = userword", &ret); //上传用户词表
if (MSP_SUCCESS != ret)
{
printf("\nMSPUploadData failed ! errorCode: %d \n", ret);
goto upload_exit;
}
upload_exit:
if (NULL != fp)
{
fclose(fp);
fp = NULL;
}
if (NULL != userwords)
{
free(userwords);
userwords = NULL;
}
return ret;
}
void run_iat(const char* audio_file, const char* session_begin_params)
{
const char* session_id = NULL;
char rec_result[BUFFER_SIZE] = {NULL};
char hints[HINTS_SIZE] = {NULL}; //hints为结束本次会话的原因描述，由用户自定义
unsigned int total_len = 0;
int aud_stat = MSP_AUDIO_SAMPLE_CONTINUE ; //音频状态
int ep_stat = MSP_EP_LOOKING_FOR_SPEECH; //端点检测
int rec_stat = MSP_REC_STATUS_SUCCESS ; //识别状态
int errcode = MSP_SUCCESS ;
FILE* f_pcm = NULL;
char* p_pcm = NULL;
long pcm_count = 0;
long pcm_size = 0;
long read_size = 0;
if (NULL == audio_file)
goto iat_exit;
f_pcm = fopen(audio_file, "rb");
if (NULL == f_pcm)
{
printf("\nopen [%s] failed! \n", audio_file);
goto iat_exit;
}
fseek(f_pcm, 0, SEEK_END);
pcm_size = ftell(f_pcm); //获取音频文件大小
fseek(f_pcm, 0, SEEK_SET);
p_pcm = (char *)malloc(pcm_size);
if (NULL == p_pcm)
{
printf("\nout of memory! \n");
goto iat_exit;
}
read_size = fread((void *)p_pcm, 1, pcm_size, f_pcm); //读取音频文件内容
if (read_size != pcm_size)
{
printf("\nread [%s] error!\n", audio_file);
goto iat_exit;
}
printf("\n开始语音听写 ...\n");
session_id = QISRSessionBegin(NULL, session_begin_params, &errcode); //听写不需要语法，第一个参数为NULL
if (MSP_SUCCESS != errcode)
{
printf("\nQISRSessionBegin failed! error code:%d\n", errcode);
goto iat_exit;
}
while (1)
{
unsigned int len = 10 * FRAME_LEN; // 每次写入200ms音频(16k，16bit)：1帧音频20ms，10帧=200ms。16k采样率的16位音频，一帧的大小为640Byte
int ret = 0;
if (pcm_size < 2 * len)
len = pcm_size;
if (len <= 0)
break;
aud_stat = MSP_AUDIO_SAMPLE_CONTINUE;
if (0 == pcm_count)
aud_stat = MSP_AUDIO_SAMPLE_FIRST;
printf(">");
ret = QISRAudioWrite(session_id, (const void *)&p_pcm[pcm_count], len, aud_stat, &ep_stat, &rec_stat);
if (MSP_SUCCESS != ret)
{
printf("\nQISRAudioWrite failed! error code:%d\n", ret);
goto iat_exit;
}
pcm_count += (long)len;
pcm_size -= (long)len;
if (MSP_REC_STATUS_SUCCESS == rec_stat) //已经有部分听写结果
{
const char *rslt = QISRGetResult(session_id, &rec_stat, 0, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRGetResult failed! error code: %d\n", errcode);
goto iat_exit;
}
if (NULL != rslt)
{
unsigned int rslt_len = strlen(rslt);
total_len += rslt_len;
if (total_len >= BUFFER_SIZE)
{
printf("\nno enough buffer for rec_result !\n");
goto iat_exit;
}
strncat(rec_result, rslt, rslt_len);
}
}
if (MSP_EP_AFTER_SPEECH == ep_stat)
break;
usleep(200*1000); //模拟人说话时间间隙。200ms对应10帧的音频
}
errcode = QISRAudioWrite(session_id, NULL, 0, MSP_AUDIO_SAMPLE_LAST, &ep_stat, &rec_stat);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRAudioWrite failed! error code:%d \n", errcode);
goto iat_exit;
}
while (MSP_REC_STATUS_COMPLETE != rec_stat)
{
const char *rslt = QISRGetResult(session_id, &rec_stat, 0, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRGetResult failed, error code: %d\n", errcode);
goto iat_exit;
}
if (NULL != rslt)
{
unsigned int rslt_len = strlen(rslt);
total_len += rslt_len;
if (total_len >= BUFFER_SIZE)
{
printf("\nno enough buffer for rec_result !\n");
goto iat_exit;
}
strncat(rec_result, rslt, rslt_len);
}
usleep(150*1000); //防止频繁占用CPU
}
printf("\n语音听写结束\n");
printf("=============================================================\n");
printf("%s\n",rec_result);
printf("=============================================================\n");
iat_exit:
if (NULL != f_pcm)
{
fclose(f_pcm);
f_pcm = NULL;
}
if (NULL != p_pcm)
{ free(p_pcm);
p_pcm = NULL;
}
QISRSessionEnd(session_id, hints);
}
int main(int argc, char* argv[])
{
int ret = MSP_SUCCESS;
int upload_on = 1; //是否上传用户词表
const char* login_params = "appid = 5afda937, work_dir = ."; // 登录参数，appid与msc库绑定,请勿随意改动
/*
* sub: 请求业务类型
* domain: 领域
* language: 语言
* accent: 方言
* sample_rate: 音频采样率
* result_type: 识别结果格式
* result_encoding: 结果编码格式
*
* 详细参数说明请参阅《讯飞语音云MSC--API文档》
*/
const char* session_begin_params = "sub = iat, domain = iat, language = zh_cn, accent = mandarin, sample_rate = 16000, result_type = plain, result_encoding = utf8";
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params); //第一个参数是用户名，第二个参数是密码，均传NULL即可，第三个参数是登录参数
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed , Error code %d.\n",ret);
goto exit; //登录失败，退出登录
}
printf("\n########################################################################\n");
printf("## 语音听写(iFly Auto Transform)技术能够实时地将语音转换成对应的文字。##\n");
printf("########################################################################\n\n");
printf("演示示例选择:是否上传用户词表？\n0:不使用\n1:使用\n");
scanf("%d", &upload_on);
if (upload_on)
{
printf("上传用户词表 ...\n");
ret = upload_userwords();
if (MSP_SUCCESS != ret)
goto exit;
printf("上传用户词表成功\n");
}
run_iat("wav/iflytek02.wav", session_begin_params); //iflytek02音频内容为“中美数控”；如果上传了用户词表，识别结果为：“中美速控”。
exit:
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}

204
src/openSrc/xfei_asr/src/sch_speak.c~
View File

@ -1,204 +0,0 @@
/*
* 语音语义技术能够将语音听写业务中的内容进行语义解析。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include "qisr.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#define BUFFER_SIZE 4096
#define FRAME_LEN 640
#define HINTS_SIZE 100
void run_iat(const char* audio_file, const char* session_begin_params)
{
const char* session_id = NULL;
char rec_result[BUFFER_SIZE] = {NULL};
char hints[HINTS_SIZE] = {NULL}; //hints为结束本次会话的原因描述，由用户自定义
unsigned int total_len = 0;
int aud_stat = MSP_AUDIO_SAMPLE_CONTINUE ; //音频状态
int ep_stat = MSP_EP_LOOKING_FOR_SPEECH; //端点检测
int rec_stat = MSP_REC_STATUS_SUCCESS ; //识别状态
int errcode = MSP_SUCCESS ;
FILE* f_pcm = NULL;
char* p_pcm = NULL;
long pcm_count = 0;
long pcm_size = 0;
long read_size = 0;
if (NULL == audio_file)
goto iat_exit;
f_pcm = fopen(audio_file, "rb");
if (NULL == f_pcm)
{
printf("\nopen [%s] failed! \n", audio_file);
goto iat_exit;
}
fseek(f_pcm, 0, SEEK_END);
pcm_size = ftell(f_pcm); //获取音频文件大小
fseek(f_pcm, 0, SEEK_SET);
p_pcm = (char *)malloc(pcm_size);
if (NULL == p_pcm)
{
printf("\nout of memory! \n");
goto iat_exit;
}
read_size = fread((void *)p_pcm, 1, pcm_size, f_pcm); //读取音频文件内容
if (read_size != pcm_size)
{
printf("\nread [%s] error!\n", audio_file);
goto iat_exit;
}
session_id = QISRSessionBegin(NULL, session_begin_params, &errcode); //听写不需要语法，第一个参数为NULL
if (MSP_SUCCESS != errcode)
{
printf("\nQISRSessionBegin failed! error code:%d\n", errcode);
goto iat_exit;
}
while (1)
{
unsigned int len = 10 * FRAME_LEN; // 每次写入200ms音频(16k，16bit)：1帧音频20ms，10帧=200ms。16k采样率的16位音频，一帧的大小为640Byte
int ret = 0;
if (pcm_size < 2 * len)
len = pcm_size;
if (len <= 0)
break;
aud_stat = MSP_AUDIO_SAMPLE_CONTINUE;
if (0 == pcm_count)
aud_stat = MSP_AUDIO_SAMPLE_FIRST;
printf(">");
ret = QISRAudioWrite(session_id, (const void *)&p_pcm[pcm_count], len, aud_stat, &ep_stat, &rec_stat);
if (MSP_SUCCESS != ret)
{
printf("\nQISRAudioWrite failed! error code:%d\n", ret);
goto iat_exit;
}
pcm_count += (long)len;
pcm_size -= (long)len;
if (MSP_REC_STATUS_SUCCESS == rec_stat) //已经有部分听写结果
{
const char *rslt = QISRGetResult(session_id, &rec_stat, 0, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRGetResult failed! error code: %d\n", errcode);
goto iat_exit;
}
if (NULL != rslt)
{
unsigned int rslt_len = strlen(rslt);
total_len += rslt_len;
if (total_len >= BUFFER_SIZE)
{
printf("\nno enough buffer for rec_result !\n");
goto iat_exit;
}
strncat(rec_result, rslt, rslt_len);
}
}
if (MSP_EP_AFTER_SPEECH == ep_stat)
break;
usleep(200*1000); //模拟人说话时间间隙。200ms对应10帧的音频
}
errcode = QISRAudioWrite(session_id, NULL, 0, MSP_AUDIO_SAMPLE_LAST, &ep_stat, &rec_stat);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRAudioWrite failed! error code:%d \n", errcode);
goto iat_exit;
}
while (MSP_REC_STATUS_COMPLETE != rec_stat)
{
const char *rslt = QISRGetResult(session_id, &rec_stat, 0, &errcode);
if (MSP_SUCCESS != errcode)
{
printf("\nQISRGetResult failed, error code: %d\n", errcode);
goto iat_exit;
}
if (NULL != rslt)
{
unsigned int rslt_len = strlen(rslt);
total_len += rslt_len;
if (total_len >= BUFFER_SIZE)
{
printf("\nno enough buffer for rec_result !\n");
goto iat_exit;
}
strncat(rec_result, rslt, rslt_len);
}
usleep(150*1000); //防止频繁占用CPU
}
printf("\n语音语义解析结束\n");
printf("=============================================================\n");
printf("%s\n",rec_result);
printf("=============================================================\n");
iat_exit:
if (NULL != f_pcm)
{
fclose(f_pcm);
f_pcm = NULL;
}
if (NULL != p_pcm)
{ free(p_pcm);
p_pcm = NULL;
}
QISRSessionEnd(session_id, hints);
}
int main(int argc, char* argv[])
{
int ret = MSP_SUCCESS;
int upload_on = 1; //是否上传用户词表
const char* login_params = "appid = 5afda937"; // 登录参数，appid与msc库绑定,请勿随意改动
/*
* sub: 请求业务类型
* domain: 领域
* language: 语言
* accent: 方言
* sample_rate: 音频采样率
* result_type: 识别结果格式
* result_encoding: 结果编码格式
*
* nlp_version: 语义版本
* sch: 是否使用语义标识
* 详细参数说明请参阅《iFlytek MSC Reference Manual》
*/
const char* session_begin_params ="nlp_version =2.0,sch=1,sub=iat,domain = iat, language = zh_cn, accent = mandarin,aue = ico, sample_rate = 16000, result_type = plain, result_encoding = gb2312";
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params); //第一个参数是用户名，第二个参数是密码，均传NULL即可，第三个参数是登录参数
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed , Error code %d.\n",ret);
goto exit; //登录失败，退出登录
}
run_iat("wav/weather.pcm", session_begin_params);
exit:
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}

104
src/openSrc/xfei_asr/src/sch_text.c~
View File

@ -1,104 +0,0 @@
/*
* 文本语义技术能将文本内容进行语义解析。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include "msp_cmn.h"
#include "msp_errors.h"
#define SOURCETEXT "source.txt" //语义文本资源
#define RESULTTEXT "result.txt" //语义结果文本
int main(int argc, char* argv[])
{
const char* login_params = "appid = 5afda937"; // 登录参数，appid与msc库绑定,请勿随意改动
const char* rec_text = NULL;
unsigned int str_len = 0;
int ret = MSP_SUCCESS;
FILE* fw = NULL;
FILE* fr = NULL;
long txtSize = 0;
long read_size = 0;
char* text = NULL;
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params); //第一个参数是用户名，第二个参数是密码，均传NULL即可，第三个参数是登录参数
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed , Error code %d.\n",ret);
goto exit; //登录失败，退出登录
}
fr=fopen(SOURCETEXT,"rb");
if(NULL == fr)
{
printf("\nopen [%s] failed! \n",SOURCETEXT);
goto exit;
}
fseek(fr, 0, SEEK_END);
txtSize = ftell(fr);
fseek(fr, 0, SEEK_SET);
text = (char *)malloc(txtSize+1);
if (NULL == text)
{
printf("\nout of memory! \n");
goto exit;
}
read_size = fread((void *)text,1, txtSize, fr);
if (read_size != txtSize)
{
printf("\nread [%s] error!\n", SOURCETEXT);
goto exit;
}
text[txtSize]='\0';
str_len = strlen(text);
printf("\n开始语义解析...\n");
rec_text = MSPSearch("nlp_version=2.0",text,&str_len,&ret);
if(MSP_SUCCESS !=ret)
{
printf("MSPSearch failed ,error code is:%d\n",ret);
goto exit;
}
printf("\n语义解析完成!\n");
fw=fopen(RESULTTEXT,"wb");
if(NULL == fw)
{
printf("\nopen [%s] failed! \n",RESULTTEXT);
goto exit;
}
read_size = fwrite(rec_text,1,str_len,fw);
if(read_size != str_len)
{
printf("\nwrite [%s] error!\n", RESULTTEXT);
goto exit;
}
printf("\n语义解析结果已写入%s文件\n",RESULTTEXT);
exit:
if (NULL != fr)
{
fclose(fr);
fr = NULL;
}
if (NULL != fw)
{
fclose(fw);
fw = NULL;
}
if (NULL != text)
{
free(text);
text = NULL;
}
printf("\n按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}

183
src/openSrc/xfei_asr/src/tts_sample.c~
View File

@ -1,183 +0,0 @@
/*
* 语音合成（Text To Speech，TTS）技术能够自动将任意文字实时转换为连续的
* 自然语音，是一种能够在任何时间、任何地点，向任何人提供语音信息服务的
* 高效便捷手段，非常符合信息时代海量数据、动态更新和个性化查询的需求。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include "qtts.h"
#include "msp_cmn.h"
#include "msp_errors.h"
/* wav音频头部格式 */
typedef struct _wave_pcm_hdr
{
char riff[4]; // = "RIFF"
int size_8; // = FileSize - 8
char wave[4]; // = "WAVE"
char fmt[4]; // = "fmt "
int fmt_size; // = 下一个结构体的大小 : 16
short int format_tag; // = PCM : 1
short int channels; // = 通道数 : 1
int samples_per_sec; // = 采样率 : 8000 | 6000 | 11025 | 16000
int avg_bytes_per_sec; // = 每秒字节数 : samples_per_sec * bits_per_sample / 8
short int block_align; // = 每采样点字节数 : wBitsPerSample / 8
short int bits_per_sample; // = 量化比特数: 8 | 16
char data[4]; // = "data";
int data_size; // = 纯数据长度 : FileSize - 44
} wave_pcm_hdr;
/* 默认wav音频头部数据 */
wave_pcm_hdr default_wav_hdr =
{
{ 'R', 'I', 'F', 'F' },
0,
{'W', 'A', 'V', 'E'},
{'f', 'm', 't', ' '},
16,
1,
1,
16000,
32000,
2,
16,
{'d', 'a', 't', 'a'},
0
};
/* 文本合成 */
int text_to_speech(const char* src_text, const char* des_path, const char* params)
{
int ret = -1;
FILE* fp = NULL;
const char* sessionID = NULL;
unsigned int audio_len = 0;
wave_pcm_hdr wav_hdr = default_wav_hdr;
int synth_status = MSP_TTS_FLAG_STILL_HAVE_DATA;
if (NULL == src_text || NULL == des_path)
{
printf("params is error!\n");
return ret;
}
fp = fopen(des_path, "wb");
if (NULL == fp)
{
printf("open %s error.\n", des_path);
return ret;
}
/* 开始合成 */
sessionID = QTTSSessionBegin(params, &ret);
if (MSP_SUCCESS != ret)
{
printf("QTTSSessionBegin failed, error code: %d.\n", ret);
fclose(fp);
return ret;
}
ret = QTTSTextPut(sessionID, src_text, (unsigned int)strlen(src_text), NULL);
if (MSP_SUCCESS != ret)
{
printf("QTTSTextPut failed, error code: %d.\n",ret);
QTTSSessionEnd(sessionID, "TextPutError");
fclose(fp);
return ret;
}
printf("正在合成 ...\n");
fwrite(&wav_hdr, sizeof(wav_hdr) ,1, fp); //添加wav音频头，使用采样率为16000
while (1)
{
/* 获取合成音频 */
const void* data = QTTSAudioGet(sessionID, &audio_len, &synth_status, &ret);
if (MSP_SUCCESS != ret)
break;
if (NULL != data)
{
fwrite(data, audio_len, 1, fp);
wav_hdr.data_size += audio_len; //计算data_size大小
}
if (MSP_TTS_FLAG_DATA_END == synth_status)
break;
printf(">");
usleep(150*1000); //防止频繁占用CPU
}//合成状态synth_status取值请参阅《讯飞语音云API文档》
printf("\n");
if (MSP_SUCCESS != ret)
{
printf("QTTSAudioGet failed, error code: %d.\n",ret);
QTTSSessionEnd(sessionID, "AudioGetError");
fclose(fp);
return ret;
}
/* 修正wav文件头数据的大小 */
wav_hdr.size_8 += wav_hdr.data_size + (sizeof(wav_hdr) - 8);
/* 将修正过的数据写回文件头部,音频文件为wav格式 */
fseek(fp, 4, 0);
fwrite(&wav_hdr.size_8,sizeof(wav_hdr.size_8), 1, fp); //写入size_8的值
fseek(fp, 40, 0); //将文件指针偏移到存储data_size值的位置
fwrite(&wav_hdr.data_size,sizeof(wav_hdr.data_size), 1, fp); //写入data_size的值
fclose(fp);
fp = NULL;
/* 合成完毕 */
ret = QTTSSessionEnd(sessionID, "Normal");
if (MSP_SUCCESS != ret)
{
printf("QTTSSessionEnd failed, error code: %d.\n",ret);
}
return ret;
}
int main(int argc, char* argv[])
{
int ret = MSP_SUCCESS;
const char* login_params = "appid = 5afda937, work_dir = .";//登录参数,appid与msc库绑定,请勿随意改动
/*
* rdn: 合成音频数字发音方式
* volume: 合成音频的音量
* pitch: 合成音频的音调
* speed: 合成音频对应的语速
* voice_name: 合成发音人
* sample_rate: 合成音频采样率
* text_encoding: 合成文本编码格式
*
* 详细参数说明请参阅《讯飞语音云MSC--API文档》
*/
const char* session_begin_params = "voice_name = xiaoyan, text_encoding = utf8, sample_rate = 16000, speed = 50, volume = 50, pitch = 50, rdn = 2";
const char* filename = "tts_sample.wav"; //合成的语音文件名称
const char* text = "亲爱的用户，您好，这是一个语音合成示例，感谢您对科大讯飞语音技术的支持！科大讯飞是亚太地区最大的语音上市公司，股票代码：002230"; //合成文本
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params);//第一个参数是用户名，第二个参数是密码，第三个参数是登录参数，用户名和密码可在http://open.voicecloud.cn注册获取
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed, error code: %d.\n", ret);
goto exit ;//登录失败，退出登录
}
printf("\n###########################################################################\n");
printf("## 语音合成（Text To Speech，TTS）技术能够自动将任意文字实时转换为连续的 ##\n");
printf("## 自然语音，是一种能够在任何时间、任何地点，向任何人提供语音信息服务的 ##\n");
printf("## 高效便捷手段，非常符合信息时代海量数据、动态更新和个性化查询的需求。 ##\n");
printf("###########################################################################\n\n");
/* 文本合成 */
printf("开始合成 ...\n");
ret = text_to_speech(text, filename, session_begin_params);
if (MSP_SUCCESS != ret)
{
printf("text_to_speech failed, error code: %d.\n", ret);
}
printf("合成完毕\n");
exit:
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}

234
src/openSrc/xfei_asr/src/tts_subscribe_speak.cpp~
View File

@ -1,234 +0,0 @@
/*
* 语音合成（Text To Speech，TTS）技术能够自动将任意文字实时转换为连续的
* 自然语音，是一种能够在任何时间、任何地点，向任何人提供语音信息服务的
* 高效便捷手段，非常符合信息时代海量数据、动态更新和个性化查询的需求。
*/
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <unistd.h>
#include <errno.h>
#include "qtts.h"
#include "msp_cmn.h"
#include "msp_errors.h"
#include "ros/ros.h"
#include "std_msgs/String.h"
#include <sstream>
#include <sys/types.h>
#include <sys/stat.h>
/* wav音频头部格式 */
typedef struct _wave_pcm_hdr
{
char riff[4]; // = "RIFF"
int size_8; // = FileSize - 8
char wave[4]; // = "WAVE"
char fmt[4]; // = "fmt "
int fmt_size; // = 下一个结构体的大小 : 16
short int format_tag; // = PCM : 1
short int channels; // = 通道数 : 1
int samples_per_sec; // = 采样率 : 8000 | 6000 | 11025 | 16000
int avg_bytes_per_sec; // = 每秒字节数 : samples_per_sec * bits_per_sample / 8
short int block_align; // = 每采样点字节数 : wBitsPerSample / 8
short int bits_per_sample; // = 量化比特数: 8 | 16
char data[4]; // = "data";
int data_size; // = 纯数据长度 : FileSize - 44
} wave_pcm_hdr;
/* 默认wav音频头部数据 */
wave_pcm_hdr default_wav_hdr =
{
{ 'R', 'I', 'F', 'F' },
0,
{'W', 'A', 'V', 'E'},
{'f', 'm', 't', ' '},
16,
1,
1,
16000,
32000,
2,
16,
{'d', 'a', 't', 'a'},
0
};
/* 文本合成 */
int text_to_speech(const char* src_text, const char* des_path, const char* params)
{
int ret = -1;
FILE* fp = NULL;
const char* sessionID = NULL;
unsigned int audio_len = 0;
wave_pcm_hdr wav_hdr = default_wav_hdr;
int synth_status = MSP_TTS_FLAG_STILL_HAVE_DATA;
if (NULL == src_text || NULL == des_path)
{
printf("params is error!\n");
return ret;
}
fp = fopen(des_path, "wb");
if (NULL == fp)
{
printf("open %s error.\n", des_path);
return ret;
}
/* 开始合成 */
sessionID = QTTSSessionBegin(params, &ret);
if (MSP_SUCCESS != ret)
{
printf("QTTSSessionBegin failed, error code: %d.\n", ret);
fclose(fp);
return ret;
}
ret = QTTSTextPut(sessionID, src_text, (unsigned int)strlen(src_text), NULL);
if (MSP_SUCCESS != ret)
{
printf("QTTSTextPut failed, error code: %d.\n",ret);
QTTSSessionEnd(sessionID, "TextPutError");
fclose(fp);
return ret;
}
printf("正在合成 ...\n");
fwrite(&wav_hdr, sizeof(wav_hdr) ,1, fp); //添加wav音频头，使用采样率为16000
while (1)
{
/* 获取合成音频 */
const void* data = QTTSAudioGet(sessionID, &audio_len, &synth_status, &ret);
if (MSP_SUCCESS != ret)
break;
if (NULL != data)
{
fwrite(data, audio_len, 1, fp);
wav_hdr.data_size += audio_len; //计算data_size大小
}
if (MSP_TTS_FLAG_DATA_END == synth_status)
break;
printf(">");
usleep(150*1000); //防止频繁占用CPU
}//合成状态synth_status取值请参阅《讯飞语音云API文档》
printf("\n");
if (MSP_SUCCESS != ret)
{
printf("QTTSAudioGet failed, error code: %d.\n",ret);
QTTSSessionEnd(sessionID, "AudioGetError");
fclose(fp);
return ret;
}
/* 修正wav文件头数据的大小 */
wav_hdr.size_8 += wav_hdr.data_size + (sizeof(wav_hdr) - 8);
/* 将修正过的数据写回文件头部,音频文件为wav格式 */
fseek(fp, 4, 0);
fwrite(&wav_hdr.size_8,sizeof(wav_hdr.size_8), 1, fp); //写入size_8的值
fseek(fp, 40, 0); //将文件指针偏移到存储data_size值的位置
fwrite(&wav_hdr.data_size,sizeof(wav_hdr.data_size), 1, fp); //写入data_size的值
fclose(fp);
fp = NULL;
/* 合成完毕 */
ret = QTTSSessionEnd(sessionID, "Normal");
if (MSP_SUCCESS != ret)
{
printf("QTTSSessionEnd failed, error code: %d.\n",ret);
}
return ret;
}
void xfcallback(const std_msgs::String::ConstPtr& msg)
{
char cmd[2000];
const char* text;
int ret = MSP_SUCCESS;
const char* session_begin_params = "voice_name = xiaoyan, text_encoding = utf8, sample_rate = 16000, speed = 50, volume = 50, pitch = 50, rdn = 2";
const char* filename = "tts_sample.wav"; //合成的语音文件名称
std::cout<<"I heard :"<<msg->data.c_str()<<std::endl;
text = msg->data.c_str();
/* 文本合成 */
printf("开始合成 ...\n");
ret = text_to_speech(text, filename, session_begin_params);
if (MSP_SUCCESS != ret)
{
printf("text_to_speech failed, error code: %d.\n", ret);
}
printf("合成完毕\n");
unlink("/tmp/cmd");
mkfifo("/tmp/cmd", 0777);
popen("mplayer -quiet -slave -input file=/tmp/cmd 'tts_sample.wav'","r");
sleep(30);
printf("Mplayer Run Success\n");
}
void toPlay()
{
}
void toExit()
{
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
}
int main(int argc, char* argv[])
{
int ret = MSP_SUCCESS;
const char* login_params = "appid = 5afda937, work_dir = .";//登录参数,appid与msc库绑定,请勿随意改动
/*
* rdn: 合成音频数字发音方式
* volume: 合成音频的音量
* pitch: 合成音频的音调
* speed: 合成音频对应的语速
* voice_name: 合成发音人
* sample_rate: 合成音频采样率
* text_encoding: 合成文本编码格式
*
* 详细参数说明请参阅《讯飞语音云MSC--API文档》
*/
/* 用户登录 */
ret = MSPLogin(NULL, NULL, login_params);//第一个参数是用户名，第二个参数是密码，第三个参数是登录参数，用户名和密码可在http://open.voicecloud.cn注册获取
if (MSP_SUCCESS != ret)
{
printf("MSPLogin failed, error code: %d.\n", ret);
/*goto exit ;*///登录失败，退出登录
toExit();
}
printf("\n###########################################################################\n");
printf("## 语音合成（Text To Speech，TTS）技术能够自动将任意文字实时转换为连续的 ##\n");
printf("## 自然语音，是一种能够在任何时间、任何地点，向任何人提供语音信息服务的 ##\n");
printf("## 高效便捷手段，非常符合信息时代海量数据、动态更新和个性化查询的需求。 ##\n");
printf("###########################################################################\n\n");
ros::init(argc,argv,"xf_tts");
ros::NodeHandle n;
ros::Subscriber sub =n.subscribe("xfwords",1000,xfcallback);
ros::spin();
exit:
printf("按任意键退出 ...\n");
getchar();
MSPLogout(); //退出登录
return 0;
}