复杂环境下运动目标跟踪

Tracking Moving Objects in Complicated Environment

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.