一种用于运动跟踪的加窗粒子滤波新算法研究

为了提高粒子滤波算法在视频跟踪中的性能,在基本粒子滤波算法的基础上,采用窗口滤波更新粒子集合,根据对目标位置估计的情况动态更新粒子集合大小,得到一种改进的粒子滤波算法--加窗粒子滤波算法.该算法利用估计窗内的混合抽样粒子集描述后验分布,通过对估计窗内具有不同权值的粒子集依据其权值大小进行抽样,并根据当前观测值对抽取的粒子状态进行更新,实现对目标的跟踪.仿真实验结果表明:这种跟踪算法在不影响跟踪精度的情况下,大大减少了计算量,较好地解决了视频目标跟踪这一非线性非高斯状态在线估计问题.

Motion Tracking Based on Novel Windowing Particle Filter

Based on the general particle filter, an advanced particle filter called windowing particle filter is introduced to improve the performance of particle filter in visual tracking by using windowbased filtering method to update particle sets and by refreshing the size particle sets according to the target state estimation. In the novel algorithm, posterior distribution is represented by mixtures of particle sets inside estimation window. The number of particles respectively sampled from different particle sets is in proportion to the weight of each particle set, and particle states are improved by integrating the latest observation to realize object tracking. The simulation results show that the proposed algorithm can reduce the computational cost for certain tracking accuracy and solve the nonlinear and non-Gaussian state estimation problem of visual tracking.