Context-based ground target tracking — An integrated approach

The general focus of this paper is the improvement of state-of-the-art Bayesian tracking filters specialized to the domain of ground moving objects to obtain high-quality track information, based on airborne ground moving target indication (GMTI) radar measurements. To counteract the numerous challenges, in particular, imprecise measurements and missed detections, a strong false alarm background, closely-spaced targets, technical and terrain obscuration as well as complex target motion, it is highly advisable to exploit additional context information in a tracking system. Three different classes of information are considered and used as extensions of standard tracking algorithms. These are the knowledge on range and Doppler blind zones of the GMTI sensor, road network information and signal strength measurements, where the latter is used to obtain estimates of a target's mean radar cross section (RCS). The performance of differently augmented cardinalized probability hypothesis density (CPHD) filter variants is assessed based on a multi-target simulation scenario.