A Bayesian estimation for single target tracking based on state mixture models

This paper presents a Bayesian algorithm for single target tracking using state mixture model theory. Compared with the existing approaches, the proposed algorithm aims at deriving the likelihood function of all measurements. Given this, an analytic Bayesian algorithm is further proposed. Moreover, under linear Gaussian assumptions on the dynamics and measurement model, a closed-form solution is proposed. Our study demonstrates the effectiveness of the proposed method in single target detection and tracking.

► The measurement likelihood function using all the measurements is derived.
► An analytic Bayesian filter for single target tracking is proposed.
► Under linear Gaussian assumptions on the dynamic and measurement models, a closed-form solution is proposed.