New data association method for automotive radar tracking

Abstract

An improved multi-target data association method for automotive radar is presented. The new approach is formulated using the association probabilities of the joint probabilistic data association (JPDA) filter. An optimal track-to-measurement data association is accomplished using the decision logic algorithm. Simulation results in heavily cluttered conditions show that the tracking accuracy of the proposed method is about 19% better than that of the JPDA filter.