基于CPHD的双层粒子滤波目标跟踪算法

常规基于势概率假设密度滤波(Cardinalized Probability Hypothesis Density,CPHD)的粒子滤波(Particle Fil? ter,PF)跟踪算法应用于多目标跟踪时,容易遇到因粒子数量增加而带来的运算效率下降、目标数目估计不准的问题。文章基于常规粒子滤波 CPHD跟踪算法,通过部署双层粒子,提出基于势概率假设密度滤波的双层粒子滤波 (Two-Layer Particle Filter-CPHD,TLPF-CPHD)算法,以便提高目标数目及状态估计精度。仿真实验结果证明,相比于常规 PF-CPHD算法,新算法具有更好的目标数目和状态估计准确性。

Two-Layer Particle Filter Tracking Algorithm Based on CPHD

When the conventional particle filter (PF) tracking algorithm based on cardinalized probability hypothesis densi?ty (CPHD) is applied to the multi-target tracking, it is easy to encounter the problems of decreased computational efficien?cy and inaccurate estimation of the target number due to the increase in the number of particles. In this paper, based onthe conventional PF-CPHD tracking algorithm, TLPF-CPHD algorithm is proposed to improve the number of targets andthe accuracy of state estimation. The simulation results show that compared with the conventional particle filter tracking al?gorithm based on CPHD, the new algorithm has significant performance advantages in target number and state estimation accuracy.