Moving Object Detection and Tracking in Video Compressed Domain
Recently, most surveillance systems are constructed by using digital video recorder (DVR). Development of video compression technologies such as H.264/AVC has made it feasible to store high-definition video with less capacity. In addition to the video compression and storage capabilities, simultaneous multi-channel processing and intelligent monitoring functions are further required. In this paper, we propose a moving object detection method based on extraction and analysis of motion vectors in H.264/AVC compressed bit-stream. We extract motion vectors of which directions and magnitudes are less than given thresholds. Then, we identify moving objects by using labeling method, and use the information to track moving objects. From the experimental results performed on H.264/AVC compressed bitstreams, the proposed method is shown to detect and to track moving objects in compressed domain quite accurately. The proposed method can be used for network based video surveillance systems with reduced computational complexity for moving object detection and tracking.
object detection tracking H.264 compressed domain
Bong-ryul Lee Jong-doo Lee Myeong-jin Lee
School of Electronics, Telecommunications, and Computer Engineering Korea Aerospace University Goyane, Gyeoneei, 412-791, Korea
国际会议
2011 International Conference on Information and Industrial Electronics(2011年信息与工业电子国际会议 ICIIE 2011)
成都
英文
785-789
2011-01-14（万方平台首次上网日期，不代表论文的发表时间）