基于均值漂移和粒子滤波的红外目标跟踪

为了提高红外目标跟踪的准确性和稳健性,提出了基于均值漂移(mean shift)和粒子滤波(PF)相结合的红外目标跟踪方法.在PF理论框架下,使用均值漂移为一种迭代模式寻找过程,对随机粒子样本进行重新分配,使粒子向目标状态的最大后验核密度估计方向移动,在均值漂移迭代过程中对样本权值进行更新.红外目标的状态后验概率分布用重新分配的加权随机样本集表示,对随机样本集使用PF算法实现红外目标运动的跟踪.实验结果表明,和一般PF和均值漂移相比,本文方法具有优越性和更强的稳健性.

IR target tracking based on mean shift and particle filter

A novel method for infrared target tracking which combines mean shift and particle filter was proposed in order to improve the tracking accuracy and robustness. Based on the particle filter,the mean shift was introduced as an iterative mode seeking procedure, in which particles move toward the maximal posterior kernel density estimation of target state. The weights of particle samples are updated as the mean shift iterative operating. The posterior distribution of the infrared target is approximated by a set of re-weighted samples,while the infrared target tracking is implemented by the particle filter algorithm which constructed by the sample set. Experimental results show that the proposed method is more effective and robust than the independent standard particle filter and mean shift.