一种基于灰关联分析的多目标跟踪算法  

A Multi-target Tracking Algorithm Based on Grey Relational Analysis

作　　者：孙启臣[1] 郭伟震 闫倩倩[1] 周莉[1] SUN Qichen GUO Weizhen YAN Qianqian ZHOU Li(School of Information and Electrical Engineering, Ludong University, Yantai 264039, Chin)

机构地区：[1]鲁东大学信息与电气工程学院,山东烟台264039

出　　处：《鲁东大学学报（自然科学版）》2017年第1期20-25,F0003,共7页Journal of Ludong University:Natural Science Edition

基　　金：国家自然科学基金(61273152);国家自然科学基金青年项目(61304052)

摘　　要：针对多特征信息融合算法进行研究,给出了一种基于灰关联分析的多目标跟踪算法.该算法首先利用灰关联算法对多特征信息进行处理,然后利用证据距离法赋予各种特征信息不同的权重,进而利用D-S证据组合规则对多特征信息进行加权融合.将基于灰关联证据距离(GED)融合多特征信息的新算法与已有算法进行仿真对比,结果表明,本文所提新算法不仅具有较准确的目标跟踪精度,而且其时间花费较少.Multi-feature information fusion algorithm is studied and a multi-target tracking algorithm based on grey relational analysis was proposed. Firstly, multi-feature information is processed by grey relational algorithm. Secondly, different weight of diversified characteristic information was assigned by using evidence distance. Finally, multi-feature information is given weighted fusion by use of D-S evidence combination rule. In comparison with the simulation results of existing algorithms, proposed the new algorithm for feature information based on grey relational evidence distance （GED） in this paper not only has target tracking precision ,but also costs less time. fusing multi- more accurate

关 键 词：多特征信息融合 灰关联证据距离 熵权法 D-S证据组合规则 

分 类 号：G202[文化科学—传播学]