Multi target tracking with acoustic power measurements using emitted power density

This paper presents a method to achieve multi target tracking using acoustic power measurements obtained from an acoustic sensor network. We first present a novel concept called emitted power density (EPD) which is an aggregate information state that holds the emitted power distribution of all targets in the scene over the target state space. It is possible to find prediction and measurement update formulas for an EPD which is conceptually similar to a probability hypothesis density (PHD). We propose a Gaussian process based representation for making the related EPD updates using Kalman filter formulas. These updates constitute a recursive EPD-filter which is based on the discretization of the position component of the target state space. The results are illustrated on a real data scenario where experiments are done with two targets constrained to a road segment