Occlusion Detection and Tracking for Multi-video-targets Based on Bhattacharyya Coefficient and Kalman Filter Innovation
This paper introduces an alternative approach for tracking multi-target in video sequence.Based on the authors previous work,instead of using the overlap coefficient (OC),in this paper the method of combining Bhattacharyya coefficient and Kalman filter innovation term is proposed to act as the criteria for jointly detecting the occlusion occurrence.Segmentation of target is also introduced in order to closely monitor the occlusion process.When the preset criteria are not met,it implies the post-occlusion separation of the targets of interest,at which point the normal tracking technique can thus be resumed.The presented approach is put to test suggesting its robustness and feasibility in tracking multi-target in complex background scenes.
multi-target tracking Kalman filter innovation Bhattacharyya coefficient occlusion
Ken Chen Xuemei Zhao Jiaming He Chul Gyu Jhun
College of Information Science and Engineering Ningbo University Ningbo,Zhejiang,315211 China New IT Center for Education Hoseo University Asan city,Chungnam,336-795 S.Korea
国际会议
太原
英文
29-33
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）