Interactive clutter measurement density estimator for multitarget data association

Abstract

Clutter measurement density (CMD) is an environmental parameter required for target tracking algorithms, but it is a priori unknown and as such needs to be estimated for tracking accuracy. The authors propose an interactive CMD estimator which is based on a Gaussian mixture probability hypothesis density filter. The proposed algorithm estimates the intensity of the clutter generators which generate the clutter measurements in the surveillance space. Since the estimated intensity is evaluated from the set of the components that consists the kinematic information for the clutter generators, the proposed algorithm can be applied to the dynamic cluttered environments. Furthermore, the proposed algorithm utilises the clutter measurement probabilities calculated from the target tracker to eliminate the target originated measurements for CMD estimation. An improved multitarget tracking performance of the proposed algorithm is verified by a Monte Carlo simulation study.