Multi-target tracking with merged measurements using labelled random finite sets

Abstract

In real world multi-target tracking problems, the presence of merged measurements is a frequently occurring phenomenon, however, the vast majority of tracking algorithms in the literature assume that each target generates independent measurements. Allowing for the possibility of measurement merging increases the computational complexity of the multi-target tracking problem, and limited computing power has been a major factor in the dominance of algorithms that assume independent measurements. In the presence of merged measurements, these algorithms suffer from performance degradation, usually due to premature track termination. In this paper, we develop a principled Bayesian solution to this problem based on the theory of random finite sets (RFS), and a tractable implementation based on the recently proposed generalised labelled multi-Bernoulli (GLMB) filter. The performance of the proposed technique is demonstrated by simulation of a multi-target bearings-only tracking scenario, where measurements become merged due to finite resolution effects.