#include
// various tracking parameters (in seconds) const double MHI_DURATION = 0.5; const double MAX_TIME_DELTA = 0.5; const double MIN_TIME_DELTA = 0.05; const int N = 3;
//
const int CONTOUR_MAX_AERA = 16; // ring image buffer IplImage **buf = 0; int last = 0;
// temporary images
IplImage *mhi = 0; // MHI: motion history image CvFilter filter = CV_GAUSSIAN_5x5; CvConnectedComp *cur_comp, min_comp; CvConnectedComp comp; CvMemStorage *storage; CvPoint pt[4];
// 参数：
// img – 输入视频帧 // dst – 检测结果
void update_mhi( IplImage* img, IplImage* dst, int diff_threshold ) {
double timestamp = clock()/100.; // get current time in seconds CvSize size = cvSize(img->width,img->height); // get current frame size
int i, j, idx1, idx2; IplImage* silh; uchar val; float temp;
IplImage* pyr = cvCreateImage( cvSize((size.width & -2)/2, (size.height & -2)/2), 8, 1 ); CvMemStorage *stor;
CvSeq *cont, *result, *squares; CvSeqReader reader;
if( !mhi || mhi->width != size.width || mhi->height != size.height ) {
if( buf == 0 ) {
buf = (IplImage**)malloc(N*sizeof(buf[0])); memset( buf, 0, N*sizeof(buf[0])); }
for( i = 0; i < N; i++ ) {
cvReleaseImage( &buf[i] );
buf[i] = cvCreateImage( size, IPL_DEPTH_8U, 1 ); cvZero( buf[i] ); }
cvReleaseImage( &mhi );
mhi = cvCreateImage( size, IPL_DEPTH_32F, 1 ); cvZero( mhi ); // clear MHI at the beginning } // end of if(mhi)
cvCvtColor( img, buf[last], CV_BGR2GRAY ); // convert frame to grayscale
idx1 = last;
idx2 = (last + 1) % N; // index of (last - (N-1))th frame last = idx2;
// 做帧差
silh = buf[idx2];
cvAbsDiff( buf[idx1], buf[idx2], silh ); // get difference between frames
// 对差图像做二值化
cvThreshold( silh, silh, 30, 255, CV_THRESH_BINARY ); // and threshold it
cvUpdateMotionHistory( silh, mhi, timestamp, MHI_DURATION ); // update MHI
cvCvtScale( mhi, dst, 255./MHI_DURATION,
(MHI_DURATION - timestamp)*255./MHI_DURATION ); cvCvtScale( mhi, dst, 255./MHI_DURATION, 0 );
// 中值滤波，消除小的噪声
cvSmooth( dst, dst, CV_MEDIAN, 3, 0, 0, 0 );
// 向下采样，去掉噪声
cvPyrDown( dst, pyr, 7 );
cvDilate( pyr, pyr, 0, 1 ); // 做膨胀操作，消除目标的不连续空洞 cvPyrUp( pyr, dst, 7 ); //
// 下面的程序段用来找到轮廓 //
// Create dynamic structure and sequence. stor = cvCreateMemStorage(0);
cont = cvCreateSeq(CV_SEQ_ELTYPE_POINT, sizeof(CvSeq), sizeof(CvPoint) , stor);
// 找到所有轮廓
cvFindContours( dst, stor, &cont, sizeof(CvContour), CV_RETR_LIST, CV_CHAIN_APPROX_SIMPLE, cvPoint(0,0)); /*
for(;cont;cont = cont->h_next) {
// Number point must be more than or equal to 6 (for cvFitEllipse_32f).
if( cont->total < 6 ) continue;
// Draw current contour.
cvDrawContours(img,cont,CV_RGB(255,0,0),CV_RGB(255,0,0),0,1, 8, cvPoint(0,0));
} // end of for-loop: \*/
// 直接使用CONTOUR中的矩形来画轮廓 for(;cont;cont = cont->h_next) {
CvRect r = ((CvContour*)cont)->rect;
if(r.height * r.width > CONTOUR_MAX_AERA) // 面积小的方形抛弃掉
{
cvRectangle( img, cvPoint(r.x,r.y),
cvPoint(r.x + r.width, r.y + r.height), CV_RGB(255,0,0), 1, CV_AA,0); } }
// free memory
cvReleaseMemStorage(&stor); cvReleaseImage( &pyr ); }
int main(int argc, char** argv) {
IplImage* motion = 0;
CvCapture* capture = 0; //视频获取结构
if( argc == 1 || (argc == 2 && strlen(argv[1]) == 1 && isdigit(argv[1][0])))
//原型：extern int isdigit(char c);
//用法：#include
capture = cvCaptureFromCAM( argc == 2 ? argv[1][0] - '0' : 0 ); else if( argc == 2 )
capture = cvCaptureFromAVI( argv[1] ); if( capture ) {
cvNamedWindow( \ for(;;) {
IplImage* image;
if( !cvGrabFrame( capture )) //从摄像头或者视频文件中抓取帧
break; image = cvRetrieveFrame( capture ); //取回由函数cvGrabFrame抓取的图像,返回由函数cvGrabFrame 抓取的图像的指针 if( image ) {
if( !motion ) {
motion =
cvCreateImage( cvSize(image->width,image->height), 8, 1 ); cvZero( motion );
motion->origin = image->origin; ///* 0 - 顶—左结构, 1 - 底—左结构 (Windows bitmaps 风格) */ } }
update_mhi( image, motion, 60 ); cvShowImage( \ if( cvWaitKey(10) >= 0 ) break; }
cvReleaseCapture( &capture ); cvDestroyWindow( \
}
return 0; }
采用金字塔方法进行图像分割
图像分割指的是将数字图像细分为多个图像子区域的过程，在OpenCv中实现了三种跟图像分割相关的算法，它们分别是：分水岭分割算法、金字塔分割算法以及均值漂移分割算法。
分水岭分割算法
分水岭分割算法需要您或者先前算法提供标记，该标记用于指定哪些大致区域是目标，哪些大致区域是背景等等；分水岭分割算法的分割效果严重依赖于提供的标记。OpenCv中的函数cvWatershed实现了该算法
金字塔分割算法
金字塔分割算法由cvPrySegmentation所实现，该函数的使用很简单；需要注意的是图像的尺寸以及金字塔的层数，图像的宽度和高度必须能被2整除，能够被2整除的次数决定了金字塔的最大层数
均值漂移分割算法
均值漂移分割算法由cvPryMeanShiftFiltering所实现，均值漂移分割的金字塔层数只能介于[1,7]之间
友情链接一下，个人感觉比较好的这方面博客：
http://www.cnblogs.com/xrwang/archive/2010/02/28/ImageSegmentation.html 效果图：