Study of Multi-target Tracking and Data Association Based on Sequential Monte Carlo Algorithm

Abstract

A new method based on sequential Monte Carlo algorithm is proposed for tracking multi-target and data association in non-linear system. The algorithm partitions the problem of multi-target tracking into two problems: single target tracking and data association. Single target tracking is implemented by using UKF and data association by using sequential Monte Carlo algorithm. Since Particle Filter has advantages in non-linear non-Gauss system, the proposed method performs well in the experiment.