Multiple group target tracking with evolving networks and labeled box particle PHD filter
Xuan Cheng, Liping Song, Zhibin Zou · 2018
This paper proposes a novel multiple group target tracking algorithm based on evolving networks and Labeled Box Particle Probability Hypothesis Density (LBP-PHD) filter. Firstly, it uses evolving networks to build group dynamic model. Secondly, LBP-PHD filter is proposed to estimate the number of targets and the targets state set. Adding different label to different box particle, it has ability to differentiate different tracks. Next, it uses established group dynamic models to update the group structure and estimate the number and the state set of group. Finally, it feeds group information back to the LBP-PHD filter's iterative process. The simulation results show that the proposed algorithm has a better performance in estimating precision and maintaining trajectory.