A randomized sampling based approach to multi-object tracking

In this paper, we present a randomized version of the finite set statistics (FISST) Bayesian recursions for multi-object tracking problems with application to the space situational awareness (SSA) problem. We introduce a hypothesis level derivation of the FISST equations that shows that the multi-object tracking problem may be considered as a finite state space Bayesian filtering problem, albeit with a growing state space. It further allows us to propose a randomized scheme, termed randomized FISST (R-FISST), where we choose the highly likely children hypotheses using Markov Chain Monte Carlo (MCMC) methods which allows us to keep the problem computationally tractable. We test the R-FISST technique on a fifty object birth and death SSA tracking and detection problem.