基于GM-PHD滤波的空间邻近多目标跟踪算法

针对传统的高斯混合概率假设密度(Gaussian mixture probability hypothesis density,GM-PHD)滤波器在跟踪空间邻近目标时存在错误估计、虚警和漏警问题,本文提出了一种改进算法.首先,提出一种权值重分配方案,对目标的高斯分量权值进行重分配,以提高目标邻近时GM-PHD滤波器的...

Multi-target tracking algorithm based on GM-PHD filter for spatially close targets

Considering the problem of wrong estimate, missing alarm and false alarm when the Gaussian mixture probability hypothesis density (GM-PHD) filter is used to track targets which are spatially close, an improved algorithm is proposed. Firstly, by arranging the weights of Gaussian components assigned to each target, a weight rearrangement scheme is proposed to improve the tracking accuracy of the GM-PHD filter when targets are spatially close. Then, based on continuous property of the target trajectory, the missed target at the current time is refined by the predicted value at the last time to reduce the missing alarm. Finally, the estimated targets are classified by making full use of the multi-frame estimated target states, and the false alarm is detected and deleted. Simulation results demonstrate that the improved algorithm has a better tracking performance compared with the existing algorithms.