C++ opencv颜色检测内存泄漏

我试图创建一个程序，可以检测三种颜色的运动。 我只用一种颜色就做得很好

但是，当我再添加两个时，当我查看任务管理器时，程序会变得滞后，占用大量内存

调试之后，我发现了一段导致内存泄漏的代码

void getMarkerPosition(IplImage* frame)
{
    //create a blank image and assigned to 'imgTracking' which has the same size of original video
    imgTracking=cvCreateImage(cvGetSize(frame),IPL_DEPTH_8U, 3);
    cvZero(imgTracking); //covert the image, 'imgTracking' to black

    IplImage* imgHSV2 = NULL; 
        imgHSV2 = cvCreateImage(cvGetSize(frame), IPL_DEPTH_8U, 3);
        cvCvtColor(frame, imgHSV2, CV_BGR2HSV); //Change the color format from BGR to HSV
        IplImage* imgThresh2 = NULL;
        imgThresh2 =  GetMarker2(imgHSV2);
        trackMarker(imgThresh2, &startLineX, &startLineY);

        imgThresh2 = GetMarker(imgHSV2);
        trackMarker(imgThresh2, &endLineX, &endLineY);

        startSection1=startLineX;
        endSection1=startSection1+73;

        startSection2=endSection1+6;
        endSection2=startSection2+73;

        startSection3=endSection2+6;
        endSection3=startSection3+73;

        startSection4=endSection3+6;
        endSection4=startSection4+73;

        startSection5=endSection4+6;
        endSection5=startSection5+73;

        startSection6=endSection5+6;
        endSection6=startSection6+73;

        startSection7=endSection6+6;
        endSection7=startSection7+73;

        cvLine(imgTracking,cv::Point(startLineX,startLineY-40) , cv::Point(endLineX,endLineY-40),cv::Scalar(0,200,0),2,8);

        cvLine(imgTracking,cv::Point(startSection1,startLineY+20) , cv::Point(endSection1,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection2,startLineY+20) , cv::Point(endSection2,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection3,startLineY+20) , cv::Point(endSection3,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection4,startLineY+20) , cv::Point(endSection4,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection5,startLineY+20) , cv::Point(endSection5,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection6,startLineY+20) , cv::Point(endSection6,endLineY+20),cv::Scalar(200,0,0),2,8);
        cvLine(imgTracking,cv::Point(startSection7,startLineY+20) , cv::Point(endSection7,endLineY+20),cv::Scalar(200,0,0),2,8);

        // Add the tracking image and the frame


cvAdd(frame, imgTracking, frame);

}*/

但我的问题是，我不知道是哪个导致了内存泄漏。请帮我找到它在哪里

---

谢谢你的关注。我等待您的任何回复。

如果您使用的是不推荐使用的c-api，您必须为我拥有的每个CVCreateImages调用cvReleaseImage，我将在我的main方法中释放该映像。我应该释放此功能中的图像吗？是！停止使用全局变量！有很好的理由，为什么他们切换到C++／CV：：Mat和Suk.Err，你知道我能参考的任何表格吗？我不知道OpenCV C函数的OpenCV C++等价物