1. 工程.pro 配置:
QT += multimedia texttospeech network
2.用windows上可用的麦克风进行录音:
audio.h 文件:
#ifndef AUDIO_H
#define AUDIO_H
#include <QObject>
#include <QAudioDeviceInfo> //查询音频设备
#include <QAudioInput> //音频输入
#include <QAudioFormat> //音频存储
#include <QFile>
#include <speech.h>
#include <QMessageBox>
#include <QDebug>
class Audio : public QObject
{
Q_OBJECT
public:
explicit Audio(QObject *parent = nullptr);
signals:
public slots:
public:
void startAudio(QString fileName);//初始化
void stopAudio();//停止录音
QString startSpeech();//开始录音并识别
private:
QString AudioFileName;//用于记录音频文件名
QFile *AudioFile;//用于操作音频文件
QAudioInput *AudioDevice;//音频设备对象
Speech *speech;//语音识别对象
};
#endif // AUDIO_H
audio.cpp文件:
#include "audio.h"
Audio::Audio(QObject *parent) : QObject(parent)
{
speech = new Speech(this);
}
void Audio::startAudio(QString fileName)
{
if(fileName.isEmpty())
{
QMessageBox::warning(NULL,"警告(Audio)","无音频文件名");
return;
}
QAudioDeviceInfo device = QAudioDeviceInfo::defaultInputDevice();//用于检测音频设备状态信息
if(device.isNull())
{
QMessageBox::warning(NULL,"警告(Audio)","无音频设备");
}else{
/* 记录录音文件 */
AudioFileName = fileName;
/* 音频编码要求 */
QAudioFormat m_format;
/* 设置采样频率 */
m_format.setSampleRate(16000);
/* 设置通道数 */
m_format.setChannelCount(1);
/* 设置位深 */
m_format.setSampleSize(16);
/* 设置编码格式 */
m_format.setCodec("audio/pcm");//http上传推荐pcm格式;也可封装成json上传
/* 判断设备是否支持该格式 */
if(!device.isFormatSupported(m_format))
{
/* 寻找最接近的格式 */
m_format = device.nearestFormat(m_format);
}
//打开文件,创建一个音频文件
AudioFile = new QFile;
AudioFile->setFileName(fileName);
AudioFile->open(QIODevice::WriteOnly);
/* 创建录音对象 */
AudioDevice = new QAudioInput(m_format,this);
AudioDevice->start(AudioFile);
}
}
QString Audio::startSpeech()
{
if(AudioFileName.isEmpty())//检查录音文件是否存在
{
QMessageBox::warning(NULL,"警告","文件不存在");
return QString("");
}
return speech->speechIdentify(AudioFileName); //语音识别,并返回识别后的结果,调用speech中的函数
}
void Audio::stopAudio()
{
/* 停止录音 */
AudioDevice->stop();
/* 关闭文件 */
AudioFile->close();
/* 删除文件对象指针并置空 */
delete AudioFile;
AudioFile = nullptr;
}
3. 通过把录音文件 通过http请求百度云语音库的方式 ,进行语音转文字
http.h文件:
#ifndef HTTP_H
#define HTTP_H
#include <QObject>
#include <QNetworkAccessManager> //发送请求
#include <QNetworkRequest> //请求内容
#include <QNetworkReply> //返回的结果
#include <QEventLoop>
class Http : public QObject
{
Q_OBJECT
public:
explicit Http(QObject *parent = nullptr);
signals:
public slots:
static bool http_postRequst(QString Url, QMap<QString,QString>header, QByteArray &requestData, QByteArray &replyData);
};
#endif // HTTP_H
http.cpp文件:
#include "http.h"
Http::Http(QObject *parent) : QObject(parent)
{
}
bool Http::http_postRequst(QString Url, QMap<QString, QString> header, QByteArray &requestData, QByteArray &replyData)
{
QNetworkAccessManager manager; //请求者
QNetworkRequest request; //请求内容
request.setUrl(Url); //获取token值时,参数只需要url和接收token值即可
QMapIterator<QString, QString> it(header);
while (it.hasNext()) {//判断header中是否有内容,并读取
it.next();
request.setRawHeader(it.key().toLatin1(),it.value().toLatin1());
}
/* 发送请求等待响应 */
QNetworkReply *Reply = manager.post(request,requestData);//发送请求
QEventLoop l;
connect(Reply,&QNetworkReply::finished,&l,&QEventLoop::quit); //等待响应
l.exec();//阻塞等待响应
if(Reply != nullptr && Reply->error() == QNetworkReply::NoError)
{
replyData = Reply->readAll();//读取http请求后的结果
return true;
}
return false;
}
请求语句的实现与replyData的解析 :
speech.h文件:
#ifndef SPEECH_H
#define SPEECH_H
#include <QObject>
#include "http.h"
#include <QFile>
#include <QMap>
#include <QByteArray>
#include <QMessageBox>
#include <QJsonDocument>
#include <QJsonObject>
#include <QJsonArray>
#include <QJsonValue>
#include <QHostInfo>
#include <QDebug>
#include <QMessageBox>
const QString BaiduSpeechUrl = "https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id=%1&client_secret=%2"; //获取token请求的Url,注册百度语音识别可获得,免费180天
const QString BaiduSpeechClientID = "5VAJlZ5OMInnkkfzLlaWDktx"; //AK
const QString BaiduSpeechClientSecret = "JWNha3VuBFpotrXBag2Tf3ALWUCOwFuO"; //SK
//语音识别的Url
//const QString BaiduSpeechSatrtUrl = "https://vop.baidu.com/server_api?dev_pid=80001&cuid=%1&token=%2";//短语音识别极速版
const QString BaiduSpeechSatrtUrl = "https://vop.baidu.com/server_api?dev_pid=1537&cuid=%1&token=%2"; //短语音识别标准版
//80001识别普通话 %1:本机唯一标识(本机标识) %2:获取的token
class Speech : public QObject
{
Q_OBJECT
public:
explicit Speech(QObject *parent = nullptr);
signals:
public slots:
public:
QString speechIdentify(QString audioFile);
private:
QString getJsonValue(QByteArray &data, QString &key);
private:
QString accessToken;
};
#endif // SPEECH_H
speech.cpp文件:
#include "speech.h"
Speech::Speech(QObject *parent) : QObject(parent)
{
}
QString Speech::speechIdentify(QString audioFile)
{
if(audioFile.isEmpty())
{
QMessageBox::warning(NULL,"警告(speech)","文件不存在");
return QString("");
}
/* 组装access token的Url */
QString TokenUrl = QString(BaiduSpeechUrl).arg(BaiduSpeechClientID).arg(BaiduSpeechClientSecret);
/**************获取token值时不用参数,但自定义http函数需要传入**************/
QMap<QString, QString>header;
header.insert(QString("Content-Type"),QString("audio/pcm;rate=16000"));//Content-Type: audio/pcm;rate=16000键值对 PCM方式上传音频
QByteArray requestData;//用于存放上传的数据,即录音信息
QByteArray replyData;//定义存放Http请求结果的字节数组
/**************获取token值不用的参数,但自定义http_postRequst函数需要传入**************/
//获取token值,获取一次使用30天
if(accessToken.isEmpty() == true)//如果还未获取token值
{
bool ret = Http::http_postRequst(TokenUrl, header, requestData, replyData);
if(ret)
{
QString key = "access_token";
accessToken = getJsonValue(replyData,key);
replyData.clear();
qDebug() << "获取的token ——" << accessToken;//调试
}
}
/* 识别语音请求 将获取的token值组装到新的url中用于发送语音识别请求*/
QString speechUrl = QString(BaiduSpeechSatrtUrl).arg(QHostInfo::localHostName()).arg(accessToken);
qDebug()<<"speechUrl"<<speechUrl;
/* 把录音文件转换成QByteArray */
QFile file;
file.setFileName(audioFile);//获取保存的录音文件
file.open(QIODevice::ReadOnly);
requestData = file.readAll();//读取录音文件中的内容
file.close();
if(requestData.isEmpty())
{
return QString("语音数据为空");
}
/* 再次发送请求 语音识别请求 */
bool ret = Http::http_postRequst(speechUrl, header, requestData, replyData);
if(ret)
{
QString key = "result";
//解析获取的json信息为QString
qDebug()<<"hello xjh "<<replyData;
QString text = getJsonValue(replyData,key);//自定义函数,下方
return text;//返回识别后的结果
}else{
QMessageBox::warning(NULL,"警告(speech)","识别失败");
}
return QString("");
}
QString Speech::getJsonValue(QByteArray &data, QString &key) //自定义json解析函数
{
QJsonParseError parseError;//json错误判定对象
QJsonDocument jsonDocument = QJsonDocument::fromJson(data,&parseError);//转换成json文档
QString retStr = "";
if(parseError.error == QJsonParseError::NoError)//判断无误
{
if(jsonDocument.isObject())
{
/* 将jsonDocument 转换成json对象 */
QJsonObject jsonObj = jsonDocument.object();
if(jsonObj.contains(key))
{
QJsonValue jsonVal = jsonObj.value(key);
if(jsonVal.isString())
{
return jsonVal.toString();
}
else if(jsonVal.isArray())//检查是否为数组
{
QJsonArray arr = jsonVal.toArray();
for(int index = 0; index<arr.size(); index++)
{
QJsonValue subValue = arr.at(index);
if(subValue.isString())
{
retStr += subValue.toString()+" ";
}
}
return retStr;
}
}
else
{
qDebug() << "不包含关键字:" << key;
}//contains(key)
}
else
{
qDebug() << "不是json对象";
}//isObject
}
else
{
qDebug() << "未成功解析JSON";
}//NoError
qDebug() << "未成功解析JSON:" << data.data();
return QString("");
}
4.主窗体上实现申明调用:
头文件:
#include <QTextToSpeech>
#include <QVoice>
#include "audio.h"//语音识别
申明:
Audio *my_audio;
QTextToSpeech *my_say;
初始化:
my_audio = new Audio;//语音识别对象 my_say = new QTextToSpeech;//文字转语音 播放
按键实现调用:
void MainWindow::on_btnStartSpeak_clicked()
{if(speakStatus==false)//开始
{
ui->btnStartSpeak->setText("结束语音");
ui->btnStartSpeak->setStyleSheet(ui->btnStartSpeak->styleSheet()+ "QToolButton{ color: red;}");
speakStatus=true;
my_audio->startAudio("audiofile");
ui->textEditReco->setText("");
}
else
{
ui->btnStartSpeak->setText("开始语音");
ui->btnStartSpeak->setStyleSheet(ui->btnStartSpeak->styleSheet()+ "QToolButton{color: blue;}");
speakStatus=false;my_audio->stopAudio();
QString retStr = my_audio->startSpeech();//识别结果my_say->say(retStr);//播放出来
qDebug() << retStr;ui->textEditReco->setText(retStr);//显示出来
}
}
实验效果: