A Novel Method of Target Tracking Based on Tracks Splitting
Aiming at the problem of more than one measurement in a tracking gate, this paper introduces the concept of true trajectory probability based on the using of tracks splitting method, therefore the process of target tracking falls into two parts: states update and true trajectory probability update. True trajectory probability denotes the possibility of the trajectory being the track of the actual target. The updated values of the true trajectory probability depend on the prior information of measurements in current time. The method also obviates the need for track initiation and track termination. Proper assignment and dynamic adjusting of the probability gate make the real-time tracking more effective as well as reducing the consumption of the system resources.
Target tracking Tracks splitting Kalman filter Data association Bayesian probability
QUAN Hongwei PENG Dongliang XUE anke
College of Information Science and Engineering in ECUST, Shanghai, P. R. China, 200237 Hangzhou Dianzi University,Hangzhou, P. R. China, 310018
国际会议
2011 China Control and Decision Conference(2011中国控制与决策会议 CCDC)
四川绵阳
英文
742-746
2011-05-23（万方平台首次上网日期，不代表论文的发表时间）