A Bayesian Approach to Target Tracking with Finite-Set-Valued Observations

Abstract

This paper presents a Bayes recursion for tracking a target that generates multiple measurements with state dependent sensor field of view and clutter. Our Bayesian formulation is mathematically well-founded due to our use of a mathematically consistent likelihood function derived from random finite set theory. A particle implementation of the proposed filter is given. Under linear Gaussian assumptions, an exact closed form solution to the proposed recursion is derived, and efficient implementations are given.