Track Fusion in the Presence of an Interference

Abstract

The problems of poor performance and track loss in the presence of interfering targets when using a Kalman filter to fuse tracks in multiple sensor systems are addressed. Two variants of the Probabilistic Multi-Hypothesis Tracking (PMHT) algorithm for multisensor tracking are presented to address these problems. A comparison of performance is made between these two algorithms, and a heuristic algorithm based on the normalised innovations from each sensor.