OpenCV 使用光流法检测物体运动

OpenCV 可以使用光流法检测物体运动，贴上代码以及效果。

// opticalflow.cpp : 定义控制台应用程序的入口点。//#include "stdafx.h"// Example 10-1. Pyramid Lucas-Kanade optical flow code///* *************** License:**************************   Oct. 3, 2008   Right to use this code in any way you want without warrenty, support or any guarentee of it working.   BOOK: It would be nice if you cited it:   Learning OpenCV: Computer Vision with the OpenCV Library     by Gary Bradski and Adrian Kaehler     Published by O'Reilly Media, October 3, 2008    AVAILABLE AT:      http://www.amazon.com/Learning-OpenCV-Computer-Vision-Library/dp/0596516134     Or: http://oreilly.com/catalog/9780596516130/     ISBN-10: 0596516134 or: ISBN-13: 978-0596516130       OTHER OPENCV SITES:   * The source code is on sourceforge at:     http://sourceforge.net/projects/opencvlibrary/   * The OpenCV wiki page (As of Oct 1, 2008 this is down for changing over servers, but should come back):     http://opencvlibrary.sourceforge.net/   * An active user group is at:     http://tech.groups.yahoo.com/group/OpenCV/   * The minutes of weekly OpenCV development meetings are at:     http://pr.willowgarage.com/wiki/OpenCV   ************************************************** */#include <cv.h>#include <cxcore.h>#include <highgui.h>#include <stdio.h>const int MAX_CORNERS = 500;int main(int argc, char** argv) {   // Initialize, load two images from the file system, and   // allocate the images and other structures we will need for   // results. // IplImage* imgA = cvLoadImage("OpticalFlow0.jpg",CV_LOAD_IMAGE_GRAYSCALE); IplImage* imgB = cvLoadImage("OpticalFlow1.jpg",CV_LOAD_IMAGE_GRAYSCALE); CvSize      img_sz    = cvGetSize( imgA ); int         win_size = 10; IplImage* imgC = cvLoadImage("OpticalFlow1.jpg",CV_LOAD_IMAGE_UNCHANGED);  // The first thing we need to do is get the features // we want to track. // IplImage* eig_image = cvCreateImage( img_sz, IPL_DEPTH_32F, 1 ); IplImage* tmp_image = cvCreateImage( img_sz, IPL_DEPTH_32F, 1 ); int              corner_count = MAX_CORNERS; CvPoint2D32f* cornersA        = new CvPoint2D32f[ MAX_CORNERS ]; cvGoodFeaturesToTrack(  imgA,  eig_image,  tmp_image,  cornersA,  &corner_count,  0.01,  5.0,  0,  3,  0,  0.04 ); cvFindCornerSubPix(  imgA,  cornersA,  corner_count,  cvSize(win_size,win_size),  cvSize(-1,-1),  cvTermCriteria(CV_TERMCRIT_ITER|CV_TERMCRIT_EPS,20,0.03) ); // Call the Lucas Kanade algorithm // char features_found[ MAX_CORNERS ]; float feature_errors[ MAX_CORNERS ]; CvSize pyr_sz = cvSize( imgA->width+8, imgB->height/3 ); IplImage* pyrA = cvCreateImage( pyr_sz, IPL_DEPTH_32F, 1 );  IplImage* pyrB = cvCreateImage( pyr_sz, IPL_DEPTH_32F, 1 );  CvPoint2D32f* cornersB        = new CvPoint2D32f[ MAX_CORNERS ];  cvCalcOpticalFlowPyrLK(     imgA,     imgB,     pyrA,     pyrB,     cornersA,     cornersB,     corner_count,     cvSize( win_size,win_size ),     5,     features_found,     feature_errors,     cvTermCriteria( CV_TERMCRIT_ITER | CV_TERMCRIT_EPS, 20, .3 ),     0  );  // Now make some image of what we are looking at:  //  for( int i=0; i<corner_count; i++ ) {     if( features_found[i]==0|| feature_errors[i]>550 ) { //       printf("Error is %f/n",feature_errors[i]);        continue;     } //    printf("Got it/n");     CvPoint p0 = cvPoint(        cvRound( cornersA[i].x ),        cvRound( cornersA[i].y )     );     CvPoint p1 = cvPoint(        cvRound( cornersB[i].x ),        cvRound( cornersB[i].y )     );     cvLine( imgC, p0, p1, CV_RGB(255,0,0),2 );  }  cvNamedWindow("ImageA",0);  cvNamedWindow("ImageB",0);  cvNamedWindow("LKpyr_OpticalFlow",0);  cvShowImage("ImageA",imgA);  cvShowImage("ImageB",imgB);  cvShowImage("LKpyr_OpticalFlow",imgC);  cvWaitKey(0);  return 0;}

OpticalFlow0.jpg

OpticalFlow1.jpg

运行结果：