【qt+opencv】实现人脸识别打卡系统2.0

一、简介

人脸识别1.0版本：

【QT】基于人脸识别的打卡系统（QT+Opencv + SQLite）--设计过程_Jason~shen的博客-CSDN博客_qt人脸识别目录使用平台设计目标功能描述设计方案数据库设计使用说明使用平台https://blog.csdn.net/qq_40602000/article/details/99674066?spm=1001.2014.3001.5502

人脸识别2.0版本：

在粉丝们的建议下在1.0的基础上做了一定的修改

1、减少了opencv繁琐的配置，打开工程即可编译运行（解决路径配置不当编译不通过问题）；

# opencv
OPENCVPATH=$$PWD/../../3rdparty/opencv64
OPENCV_LIB=$$OPENCVPATH/x64/mingw/lib
OPENCV_INC=$$OPENCVPATH/include
INCLUDEPATH +=$$OPENCV_INC
INCLUDEPATH +=$$OPENCV_INC/opencv
INCLUDEPATH +=$$OPENCV_INC/opencv2
LIBS += $$OPENCV_LIB/libopencv*
BINPATH=$$PWD/../../bin
LIBSPATH = $$PWD/../Libs

2、去除了使用python脚本生成训练使用的at.txt文件，改为c++编写（解决没有python环境的问题）；

QStringList Util::getFileFolders(const QString &dirPath)
{
    QStringList folderList;
    QDir dir(dirPath);
    dir.setFilter(QDir::Dirs);
    foreach(QFileInfo fullDir, dir.entryInfoList())
    {
        if(fullDir.fileName() == "." || fullDir.fileName() == "..") continue;
        folderList.append(fullDir.fileName());
    }
    return folderList;
}
void Util::createAtLabel(const QString &facePath, const QString &atPath)
{
    qDebug() <<facePath;
    qDebug()<<atPath;
    deleteFileOrFolder(atPath);
    QFile file(atPath);
    file.open(QIODevice::ReadWrite | QIODevice::Text); 
    QStringList idList = getFileFolders(facePath);
    qDebug()<<idList;
    foreach (QString id, idList) {
        QString oneFacePath = facePath + "\\" + id;
        QStringList imageList = getImageFileNames(oneFacePath);
        foreach (QString name, imageList) {
            QString labelInfo = oneFacePath + "\\" + name  + ";"+id;
            file.write(labelInfo.toUtf8() + "\n"); 
        }
    }
    file.close();
}