视频监控中多摄像头下的行人跟踪.pdfVIP

Abstract Video surveillance is an essential part of the social security. With the rapid development of social economy and the increasing range of video surveillance, smart monitor becomes an urgent need. Especially, multi-camera video surveillance for the same location is very necessary in some important places. Therefore, pedestrian tracking via multi-camera has become an important issue in computer vision. Multi-camera tracking is often based on feature matching or three-dimensional reconstruction method. As information fusion from multi-camera by feature matching, when the angle difference between cameras becomes large, objects’ characteristics change a lot, the matching error rate will increase. As information fusion by three-dimensional reconstruction, it has lower error rate than the former on, but has high computational complexity and needs additional information. For these reasons, a pedestrian tracking via Multi-camera method based on homography constraint has been realized. First, the object information is needed for object tracking. By experiments, it compared variant pedestrian object extraction method and its difference when the parameters changed . Second, it calculated the homography matrix between cameras, and used homography constraint rules to fuse information from multi-camera and got a foot map. Then, it did the reverse perspective projection to get the corresponding relationship between the objects and the foot map. To improve the accuracy of the corresponding relationship, a correction strategy is proposed. It defined different confidence of tracking result from different cameras, and finally tracking