Cluster tracking under kinematical constraints using random matrices

Collectively moving object clusters are of particular interest in certain applications and have to be tracked as separate aggregated entities consisting of an unknown number of individuals. Tracking of convoys or large vehicles in wide area ground surveillance are practically important examples (GMTI: Ground Moving Target Indicator). In standard tracking algorithms, the objects of interest are usually considered as point source objects; i. e. compared to the sensor resolution their spatial extension is neglected. Due to the increasing resolution capabilities of modern sensors, however, different scattering centers of an extended object can cause distinct detections. In this sense also collectively moving object groups can be considered as extended objects. Due to the resulting data association and resolution conflicts, any attempt of tracking individual objects within the group is no longer reasonable. In this paper ellipsoidal object extensions are modeled by random matrices, which are treated as additional state variables to be estimated. An important aspect is the incorporation of context information into the Bayesian data processing formalism. We here consider kinematical constraints such as road maps and sensor specific characteristics such as Doppler-blindness.