Particle filtering and data association using attribute data

This paper presents a joint classification and multi-target tracking method using particle filters. We consider a type of identity attribute data which are easily incorporated directly into the single-target posterior expression. The classification of the target is performed with a Bayesian update of the posterior probabilities of an identity hypothesis. Also, we propose a simple method of recursively updating possible time-varying attribute data model parameters. The effectiveness of the proposed method is illustrated in a simulation study using a dense multi target scenario.