Gaussian Mixtures Match and Fusion Algorithms for Multi-Sensor Multi-Target Tracking

Abstract

The results of single sensor Gaussian Mixture PHD Tracker (GM-PHDT) will deteriorate when tracking targets in dense clutter and low detection rate environments. In this work, we propose three multi-sensor multi-target GM-PHD fusion algorithms. For the proposed algorithms, we obtain the multi-sensor posterior GMs by fusing the matched GMs from multiple sensors. Specifically, the GM matching and fusing methods are proposed to fullfill the applicable GM-PHD algorithms. The simulation results demonstrated the efficiency of the proposed algorithms