Study on moving-objects identification based on evidence theory

A novel method of moving target identification based on multi-features fusion is proposed in this paper. Firstly, the distribution model of evidence weights is set up by the improved Dempster-Shafter (D-S) algorithm in order to solve the invalidation problem of highly conflict evidences. Then applies particle swarm optimization method to get the optimum evidence weights, which modify the original basic assignment function in the condition of ensuring the minimum of whole evidence conflict. This algorithm is used for video surveillance system to distinguish vehicle, people and other objects. Through analyzing the simulation example, the results show that the algorithm can gives a more reasonable combination results and has a good adaptive ability.