Vehicle Tracking in Clutter Environment using High Resolution Radar

Vehicle tracking using automotive radar is an essential part of autonomous driving. With the resolution of radar increasing, a vehicle generates multiple measurements per time and should be considered as an extended target. By using multiple radar measurements, the tracking algorithm can estimate the kinematic state and extension of vehicle. In this paper, we propose a method to track the vehicle in clutter environment by using extended vehicle model and probabilistic data association with Markov chain Monte Carlo technique. Simulation results show the effectiveness and performance our approach for vehicle tracking under clutter environment.