详细信息
文献类型:期刊文献
中文题名:一种自适应多目标跟踪算法
英文题名:New adaptive multi-target tracking algorithm
作者:叶西宁[1];顾幸生[1];常青[1];潘泉[2]
机构:[1]华东理工大学信息工程学院,上海200237;[2]西北工业大学自动化学院,陕西西安710072
年份:2005
卷号:27
期号:11
起止页码:1858
中文期刊名:系统工程与电子技术
外文期刊名:Systems Engineering and Electronics
收录:CSTPCD;;EI(收录号:2005519605989);Scopus;北大核心:【北大核心2004】;CSCD:【CSCD2011_2012】;
基金:国家自然科学基金资助课题(60172037)
语种:中文
中文关键词:自适应算法;多目标跟踪;数据关联
外文关键词:adaptive algorithm; multi-target tracking; data association
摘要:给多目标跟踪滤波器引入一个自适应比例因子,用该比例因子描述目标运动过程中的不可预测性或机动性。并结合实际情况,针对传统的数据关联算法的可行性规则(量测与目标一一对应)所存在的问题,提出量测与目标是多-多对应的新可行性规则。在此基础上,给出一种自适应多目标跟踪算法。仿真结果表明:自适应算法的跟踪性能优于非自适应算法的跟踪性能。利用该方法对多目标进行跟踪,可以达到很好的跟踪效果。
A scale factor to multi-target tracking filter is introduced. The filter employs the scale factor to describe the target's unpredictability or maneuverability when the target moves. Furthermore, in some practical facts, the feasible rule of traditional data association methods--“ the association of measurements and tracks is one to one”, is not conformable to the practical facts. So a new feasible rule, that is “the association of measurements and tracks is multiple to multiple”, is presented. And on this base, a new adaptive multi-target tracking algorithm is given. Simulation results demonstrate that the tracking performance of the adaptive algorithm is better than that of the non-adaptive algorithm.
参考文献:
正在载入数据...
