视频运动目标跟踪，基于opencv ， vc++

//打开视频文件以及车辆跟踪和识别，按钮消息响应部分
void CTrackandIDDlg::OnStartTrackandID()
{
// TODO: Add your control notification handler code here
int argc=2;
////打开文件///////////////////////////////////////////////////
CString FilePathName;
CFileDialog dlg(TRUE);
if(dlg.DoModal()==IDOK) //
FilePathName=dlg.GetPathName();

IplImage* pFrame = NULL;
IplImage* pFrImg = NULL;
IplImage* pBkImg = NULL;
IplImage* pFrImg1 = NULL;

CvMat* pFrameMat = NULL;
CvMat* pFrMat = NULL;
CvMat* pBkMat = NULL;
CvMat* pFrMat1 = NULL;

CvMemStorage * storage = cvCreateMemStorage(0);//轮廓边缘提取时的参数
CvSeq * contour = 0;//轮廓边缘提取时的参数
int mode = CV_RETR_EXTERNAL;//轮廓边缘提取时的参数
//形态学处理时内核的大小
IplConvKernel* Element = cvCreateStructuringElementEx(13,13,1,1,CV_SHAPE_RECT,NULL);

CvFont font1;//初始化字体格式
int linetype=CV_AA;
cvInitFont(&font1, CV_FONT_HERSHEY_SIMPLEX, 0.5, 0.5, 0, 1, 8);
//用字符串时一定要把using namespace std;写在前面，否则不能用,下面是用于显示的字符串
string msg[10]={"JGD01","JGD02","JGD03","JGD04","JGD05","JGD06","JGD07","JGD08","JGD09","JGD10"};
int No=0;//用于记录显示车辆
bool FindCar=false;

//在视频中画出感兴趣的区域，怎么样才能沿车道画线？？？？？？？
CvPoint pt1,pt2,pt3,pt4,pt5;
pt1.x=292;//(视频中左下点)
pt1.y=100;
pt2.x=412;//(视频中右上点)
pt2.y=280;
CvRect bndRect=cvRect(0,0,0,0);//用cvBoundingRect画出外接矩形时需要的矩形
int avgX = 0;//The midpoint X position of the rectangle surrounding the moving objects
int avgY = 0;//The midpoint Y position of the rectangle surrounding the moving objects
int avgX1=0;//用来合并相近的车辆
int avgY1=0;
for(int i=0;i<10;i++)
{
TrackBlock[i]=NULL;
if((TrackBlock[i]=(struct AvTrackBlock *) malloc(sizeof(struct AvTrackBlock)))==NULL)
{
MessageBox("内存分配错误");
exit(1);
}
}////////////////////

CvCapture* pCapture = NULL;
int nFrmNum = 0;//表示图像的帧数

//创建窗口
cvNamedWindow("video", 1);
//cvNamedWindow("background",1);
cvNamedWindow("foreground",1);
//使窗口有序排列
cvMoveWindow("video", 30, 0);
//cvMoveWindow("background", 360, 0);
cvMoveWindow("foreground", 690, 0);

if( argc > 2 ){
fprintf(stderr, "Usage: bkgrd [video_file_name]/n");
//return -1;
}

////打开摄像头///////////////////////////////////////////////////
if (argc ==1)
if( !(pCapture = cvCaptureFromCAM(-1))){
fprintf(stderr, "Can not open camera./n");
//return -2;
}

///打开视频文件//////////////////////////////////////////////////
if(argc == 2)
if( !(pCapture = cvCaptureFromFile(FilePathName))){
fprintf(stderr, "Can not open video file %s/n", FilePathName);
//return -2;
}

//逐帧读取视频,cvQueryFrame从摄像头或者文件中抓取并返回一帧
while(pFrame = cvQueryFrame(pCapture))
{
nFrmNum++;
//如果是第一帧，需要申请内存，并初始化
if(nFrmNum == 1)
{
pBkImg = cvCreateImage(cvSize(pFrame->width, pFrame->height), IPL_DEPTH_8U,1);
pFrImg = cvCreateImage(cvSize(pFrame->width, pFrame->height), IPL_DEPTH_8U,1);

pBkMat = cvCreateMat(pFrame->height, pFrame->width, CV_32FC1);
pFrMat = cvCreateMat(pFrame->height, pFrame->width, CV_32FC1);
pFrameMat = cvCreateMat(pFrame->height, pFrame->width, CV_32FC1);
cvCvtColor(pFrame, pBkImg, CV_BGR2GRAY);
cvCvtColor(pFrame, pFrImg, CV_BGR2GRAY);
cvConvert(pFrImg, pFrameMat);
cvConvert(pFrImg, pFrMat);
cvConvert(pFrImg, pBkMat);
}

else if(nFrmNum == 3)
{
cvCvtColor(pFrame, pFrImg, CV_BGR2GRAY);
cvConvert(pFrImg, pFrameMat);
//高斯滤波先，以平滑图像
cvSmooth(pFrameMat, pFrameMat, CV_GAUSSIAN, 3, 0, 0);

//在视频中设置并画出感兴趣的区域
cvRectangle(pFrame,pt1,pt2,CV_RGB(255,0,0),2, 8, 0 );

//当前帧跟背景图相减,cvAbsDiff计算两个数组差的绝对值
cvAbsDiff(pFrameMat, pBkMat, pFrMat);

//二值化前景图
cvThreshold(pFrMat, pFrImg, 60, 255.0, CV_THRESH_BINARY);

//通过查找边界找出ROI矩形区域内的运动车辆，建立完全目标档案
//cvCanny(pFrImg, pBkImg, 50, 150, 3);
cvDilate(pFrImg,pBkImg,Element,1);
cvFindContours(pBkImg, storage, &contour, sizeof(CvContour),
mode, CV_CHAIN_APPROX_SIMPLE);
//process each moving contour in the current frame用函数cvBoundingRect
for(;contour!=0;contour=contour->h_next)
{
//Get a bounding rectangle around the moving object.
bndRect = cvBoundingRect(contour, 0);

//Get an average X position of the moving contour.
avgX = (bndRect.x + bndRect.x + bndRect.width) / 2;
avgY = (bndRect.y + bndRect.y + bndRect.height) / 2;
pt5.x = bndRect.x;//写字的左下角点
pt5.y = avgY;

//If the center of contour is within ROI than show it
if(avgX>300 && avgX<400 && avgY<300 && avgY>80)
{
pt3.x = bndRect.x;
pt3.y = bndRect.y;
pt4.x = bndRect.x + bndRect.width;
pt4.y = bndRect.y + bndRect.height;
if(bndRect.height>35) //把长度小于某个阀值的干扰矩形去掉
{
cvRectangle(pFrame,pt3,pt4,CV_RGB(255,0,0),1, 8, 0 );
//在车辆的中心写编号
cvPutText( pFrame, msg[No].c_str(), pt5, &font1, cvScalar(0,255,0));
//把当前车辆存入跟踪数组
TrackBlock[No]->Direction=1;
TrackBlock[No]->FramesTracked=nFrmNum;
TrackBlock[No]->avgX=avgX;
TrackBlock[No]->avgY=avgY;
No++;
}
}
}/////查找边界的for 循环结束

//更新背景///////////////////////////////////////////////////
cvRunningAvg(pFrameMat, pBkMat, 0.005, 0);
//将背景转化为图像格式，用以显示
cvConvert(pBkMat, pBkImg);

//显示图像////////////////////////////////////////////////////
cvShowImage("video", pFrame);
//cvShowImage("background", pBkImg);
//cvShowImage("foreground", pFrImg);

//如果有按键事件,则跳出循环,此等待也为cvShowImage函数提供时间完成显示,等待时间可以根据CPU速度调整
if( cvWaitKey(2) >= 0 )
break;
}

else if(nFrmNum > 3)//从第三帧开始，根据完全目标档案来新增或删除运动车辆档案。
{
cvCvtColor(pFrame, pFrImg, CV_BGR2GRAY);
cvConvert(pFrImg, pFrameMat);
//高斯滤波先，以平滑图像
cvSmooth(pFrameMat, pFrameMat, CV_GAUSSIAN, 3, 0, 0);

//在视频中设置并画出感兴趣的区域
//cvSetImageROI(pFrame,rect1);
cvRectangle(pFrame,pt1,pt2,CV_RGB(255,0,0),2, 8, 0 );

//当前帧跟背景图相减,cvAbsDiff计算两个数组差的绝对值
cvAbsDiff(pFrameMat, pBkMat, pFrMat);

//二值化前景图,void cvThreshold( const CvArr* src, CvArr* dst, double threshold,
//double max_value, int threshold_type );
cvThreshold(pFrMat, pFrImg, 60, 255.0, CV_THRESH_BINARY);

//通过查找边界找出ROI矩形区域内的运动车辆，建立完全目标档案
//cvCanny(pFrImg, pBkImg, 50, 150, 3);
cvDilate(pFrImg,pBkImg,Element,1);
cvFindContours( pBkImg, storage, &contour, sizeof(CvContour),
mode, CV_CHAIN_APPROX_SIMPLE);
//process each moving contour in the current frame用函数cvBoundingRect
for(;contour!=0;contour=contour->h_next)
{
//Get a bounding rectangle around the moving object.
bndRect = cvBoundingRect(contour, 0);

//Get an average X position of the moving contour.
avgX = (bndRect.x + bndRect.x + bndRect.width) / 2;
avgY = (bndRect.y + bndRect.y + bndRect.height) / 2;
pt5.x=bndRect.x;//写字的左下角点
pt5.y=avgY;

//If the center of contour is within ROI than show it
if(avgX > 300 && avgX < 400 && avgY < 280 && avgY > 100)
{
pt3.x = bndRect.x;
pt3.y = bndRect.y;
pt4.x = bndRect.x + bndRect.width;
pt4.y = bndRect.y + bndRect.height;
if(bndRect.height>35) //把长度小于某个阀值的干扰矩形去掉
{
cvRectangle(pFrame,pt3,pt4,CV_RGB(255,0,0),1, 8, 0 );
//cvPutText(pFrame,msg[No].c_str(), pt5, &font1, cvScalar(0,255,0));
//在跟踪数组中寻找看是否有匹配的车辆，没有则表示是新车辆
for(int i=0;i<10;i++)
{
if(TrackBlock[i]->avgX !=0 && abs(avgX-TrackBlock[i]->avgX)<20 &&
abs(avgY-TrackBlock[i]->avgY)<50)
{
cvPutText(pFrame,msg[i].c_str(), pt5, &font1, cvScalar(0,255,0));
TrackBlock[i]->FramesTracked=nFrmNum;
TrackBlock[i]->avgX=avgX;
TrackBlock[i]->avgY=avgY;
i=10;//使跳出for循环
FindCar=true;
}
}
if(FindCar!=true && avgY<120)//表示没有找到车辆
{
TrackBlock[No]->Direction=1;
TrackBlock[No]->FramesTracked=nFrmNum;
TrackBlock[No]->avgX=avgX;
TrackBlock[No]->avgY=avgY;
if(No==9){
No=0;
}
else
No++;
}
FindCar=false;//赋值为false为下一次寻找车辆做准备
}
}
}//轮廓分for循环结束

//对于没有匹配的车辆，表示已经出了边界，清空数组
for(int j=0;j<10;j++)
{
if(TrackBlock[j]->FramesTracked != nFrmNum)
{
//虽然置为零，但是可能零和当前中心的值在设定的范围内，所以不行。
//TrackBlock[j]=NULL;为何用NULL不行。
TrackBlock[j]->Direction=0;
TrackBlock[j]->FramesTracked=0;
TrackBlock[j]->avgX=0;
TrackBlock[j]->avgY=0;
}
}

//更新背景///////////////////////////////////////////////////
cvRunningAvg(pFrameMat, pBkMat, 0.005, 0);
//将背景转化为图像格式，用以显示
cvConvert(pBkMat, pBkImg);

//显示图像////////////////////////////////////////////////////
cvShowImage("video", pFrame);
cvShowImage("background", pBkImg);
cvShowImage("foreground", pFrImg);

/*if(nFrmNum/2 ==0)
pBkMat=pFrameMat;*/
//如果有按键事件,则跳出循环,此等待也为cvShowImage函数提供时间完成显示,等待时间可以根据CPU速度调整
if( cvWaitKey(2) >= 0 )
break;
}//
}//while循环结束

cvReleaseStructuringElement(&Element);//删除结构元素
//销毁窗口
cvDestroyWindow("video");
cvDestroyWindow("background");
cvDestroyWindow("foreground");

//释放图像和矩阵
cvReleaseImage(&pFrImg);
cvReleaseImage(&pBkImg);

cvReleaseMat(&pFrameMat);
cvReleaseMat(&pFrMat);
cvReleaseMat(&pBkMat);
}