An Efficient Track Management Scheme for the Gaussian-Mixture Probability Hypothesis Density Tracker

Abstract

The Gaussian mixture probability hypothesis density (GM-PHD) filter is a closed-form solution for the probability hypothesis density (PHD) filter, which was proposed for jointly estimating the time-varying number of targets and their states from a sequence of noisy measurement sets in the presence of data association uncertainty, clutter and miss-detections. Recently, a GM-PHD tracker based on the GM-PHD filter has been proposed to correctly maintain temporal association amongst target estimates by tagging individual Gaussian components, and to provide estimates of individual target trajectories and their identities. In this paper, we propose a tag and a track management scheme for the GM-PHD tracker, which is computationally efficient and provides a framework for parallel processing of data. Based on the proposed scheme, we also present a number of simpler and efficient pruning schemes for Gaussian components.