Open cv??? ????????? tracking
2 likes2,444 views
1. The document discusses three methods for object tracking in OpenCV: LK optical flow matching, SURF feature matching, and Kalman filtering. 2. It explains how to extract and track SURF features between frames and match features using BruteForce matching. 3. Kalman filtering is discussed as a way to predict and correct object positions frame to frame based on measurements and system dynamics.
Open cv??? ????????? tracking
- 1. LK ? ?matching SURF ?matching Kalman ?Filter OpenCV? ??? ?? ??? ?? ?? Daesik ?Jang dsjang@kunsan.ac.kr 1
- 2. 1. ?LK ???? ???? ??? ? goodFeaturesToTrack ?: ???? ??? ? calcOpBcalFlowPyrLK ?: ?LK ?????? ????? ???? ?feature-?©\ prev ?????? ?image_next?? ??? ?features_next? ??? // ?Tracker ?is ?initialised ?and ?initial ?features ?are ?stored ?in ?features_next // ?Now ?iterate ?through ?rest ?of ?images // ?Obtain ?first ?image ?and ?set ?up ?two ?feature ?vectors for(;;) cv::Mat ?image_prev, ?image_next; { std::vector<cv::Point> ?features_prev, ?features_next; ? ? ? ?image_prev ?= ?image_next.clone(); ? ? ? ?feature_prev ?= ?features_next; image_next ?= ?getImage(); ? ? ? ?image_next ?= ?getImage(); ? ?// ?Get ?next ?image // ?Obtain ?initial ?set ?of ?features ? ? ? ?// ?Find ?position ?of ?feature ?in ?new ?image cv::goodFeaturesToTrack(image_next, ?// ?the ?image ? ? ? ? ?cv::calcOpticalFlowPyrLK( ? ?features_next, ? ? ?// ?the ?output ?detected ?features ? ? ? ? ? ?image_prev, ?image_next, ?// ?2 ?consecutive ?images ? ?max_count, ? ?// ?the ?maximum ?number ?of ?features ? ? ? ? ? ? ?feature_prev, ?// ?input ?point ?positions ?in ?first ?im ? ?qlevel, ? ? ? ? ?// ?quality ?level ? ? ? ? ? ?features_next, ?// ?output ?point ?positions ?in ?the ?2nd ? ?minDist ? ? ? ? ?// ?min ?distance ?between ?two ?features ? ? ? ? ? ?status, ? ? ? ?// ?tracking ?success ); ? ? ? ? ? ?err ? ? ? ? ? ?// ?tracking ?error ? ? ? ?); ? ? ? ?if ?( ?stopTracking() ?) ?break; } 2
- 3. 2. ?SURF ???? ???? ??? ? SurfFeatureDetector ????? ?detect() ????? ???? ?SURF ? ?? ??? ¨C ??? ???? ?keypoints ???? ???? // vector of keypoints std::vector<cv::KeyPoint> keypoints; // Construct the SURF feature detector object cv::SurfFeatureDetector surf( 2500.); // threshold // Detect the SURF features surf.detect(image,keypoints); // Draw the keypoints with scale and orientation information cv::drawKeypoints(image, // original image keypoints, // vector of keypoints featureImage, // the resulting image cv::Scalar(255,255,255), cv::DrawMatchesFlags::DRAW_RICH_KEYPOINTS); //?ag 3
- 4. 2. ?SURF ???? ???? ??? 4
- 5. 2. ?SURF ???? ???? ??? ? SurfDescriptorExtractor ????? ????? ?keypoint??? ? ???? ????? ?descriptor? ??? // Construction of the SURF descriptor extractor cv::SurfDescriptorExtractor surfDesc; // Extraction of the SURF descriptors cv::Mat descriptors1; surfDesc.compute(image1,keypoints1,descriptors1); 5
- 6. 2. ?SURF ???? ???? ??? ? BruteForceMatcher ????? ????? ?descriptor? ???? ??? ? ??? ?matches? ?????. // Construction of the matcher cv::BruteForceMatcher<cv::L2<?oat>> matcher; // Match the two image descriptors std::vector<cv::DMatch> matches; matcher.match(descriptors1,descriptors2, matches); 6
- 7. 2. ?SURF ???? ???? ??? ? keypoint?? ?matching ? ???? ???? ????? ????? ? ???. cv::Mat imageMatches; cv::drawMatches( image1,keypoints1, // 1st image and its keypoints image2,keypoints2, // 2nd image and its keypoints matches, // the matches imageMatches, // the image produced cv::Scalar(255,255,255)); // color of the lines 7
- 8. 2. ?SURF ???? ???? ??? ? SURF ????? ??? ??? 8
- 9. 3. ?OpenCV? ?Kalman ?Filter? ???? ??? ? KalmanFilter ???? ???? ??? KalmanFilter ?KF(4, ?2, ?0); KF.transiBonMatrix ?= ?*(Mat_<?oat>(4, ?4) ?<< ?1,0,1,0, ? ? ?0,1,0,1, ? ?0,0,1,0, ? ?0,0,0,1); Mat_<?oat> ?measurement(2,1); ?measurement.setTo(Scalar(0)); ? // ?init... KF.statePre.at<?oat>(0) ?= ?mouse_info.x; KF.statePre.at<?oat>(1) ?= ?mouse_info.y; KF.statePre.at<?oat>(2) ?= ?0; KF.statePre.at<?oat>(3) ?= ?0; setIdenBty(KF.measurementMatrix); setIdenBty(KF.processNoiseCov, ?Scalar::all(1e-?©\4)); setIdenBty(KF.measurementNoiseCov, ?Scalar::all(1e-?©\1)); setIdenBty(KF.errorCovPost, ?Scalar::all(.1)); 9
- 10. 3. ?OpenCV? ?Kalman ?Filter? ???? ??? ? KF ???? ???? ???? ??? // ?First ?predict, ?to ?update ?the ?internal ?statePre ?variable Mat ?prediction ?= ?KF.predict(); Point ? predictPt(prediction.at<float>(0),prediction.at<float>(1)); ? ? ? ? ? ? ? ? ? ? ? ? ? // ?Get ?mouse ?point measurement(0) ?= ?mouse_info.x; measurement(1) ?= ?mouse_info.y; ? ? ? ? ? ? ? ? ? ? ? ? ? Point ?measPt(measurement(0),measurement(1)); ? // ?The ?"correct" ?phase ?that ?is ?going ?to ?use ?the ?predicted ?value ? and ?our ?measurement Mat ?estimated ?= ?KF.correct(measurement); Point ?statePt(estimated.at<float>(0),estimated.at<float>(1)); 10