Target tracking with a two-scan data association algorithm extended for the hybrid target state

Abstract

This study develops a two-scan data association algorithm for the hybrid target state, which consists of the target trajectory state and the target existence state. The proposed method uses the sliding window-based approach for the multi-scan data association. It updates the target existence state by considering all the feasible multi-scan target existence events in the sliding window. The proposed method also updates the target trajectory state by calculating the joint association events in the sliding window. Performance of the proposed algorithm is compared with that of the integrated track splitting algorithm and the integrated probabilistic data association algorithm for various conditions.