复杂环境下运动目标跟踪  

Tracking Moving Objects in Complicated Environment

作　　者：王婷婷[1] 王京玲[1] 刘剑波[1] 

机构地区：[1]中国传媒大学信息工程学院,北京100024

出　　处：《中国传媒大学学报（自然科学版）》2008年第4期24-28,共5页Journal of Communication University of China：Science and Technology

基　　金：国家自然科学基金项目(60572041)资助

摘　　要：主要针对复杂背景环境下的非刚体运动目标,采用基于核密度加权的颜色模型粒子滤波器算法对目标进行跟踪。利用巴特查里亚距离作为判据检测跟踪错误,并以此指导跟踪过程的恢复。仿真实验表明,该方法能够对复杂环境下的运动目标进行有效跟踪,并且有较强的抗干扰能力和自动恢复能力。This paper mainly works on tracking of non - rigid moving objects in complicated background environment using particle filter algorithm which is based on the color model weighted by kernel density. Use Bhattacharyya distance as the criterion to detect the target tracking error and to guide the recovery of the tracking process. The simulation experiment indicates that this algorithm can effectively track moving objects and have stronger anti - interference ability and the capability of automatic recovery.

关 键 词：目标跟踪 遮挡 粒子滤波器 颜色模型 巴特查里亚系数 

分 类 号：TN919[电子电信—通信与信息系统]