An improved multiple extended target tracking algorithm based on measurement rate estimates

Abstract

Extended target probability hypothesis density filter based on the Gaussian mixture technique, referred to as the ET-GM-PHD algorithm, has proved to be a promising algorithm for multiple extended target tracking. However, this method can only be used in the multi-target tracking systems with a known measurement rate. Otherwise, the tracking performance will decline greatly by using error value of the measurement rate. To solve this problem, an adaptive estimate method of measurement rate is proposed in this paper and which is integrated into the framework of the ET-GM-PHD filter. Moreover, the mean shift technique and the density analysis method are introduced for measurement partition. Simulation results show that the proposed algorithm can effectively estimate the unknown measurement rate and has a good performance of multiple extended target tracking with a strong robustness.