A JPDA-Based Approach for Multi-Target Tracking Using Multi-Sensor Sequential Fusion and Track Maintenance

This paper considers the multi-target tracking of unknown and time-varying number of targets using multiple sensors. The problem becomes more challenging if sensors provide inconsistent numbers of targets. To address this problem, a sequential fusion joint probabilistic data association (JPDA) algorithm with track maintenance for multi-sensor systems is developed. In particular, to enhance the track initiation method for multi-sensor scenarios, we introduce a novel clustering strategy. This strategy, combined with a sequential fusion technique, forms the basis of our newly proposed sequential track initiation algorithm. Furthermore, we integrate a cost function-based track termination algorithm into each iteration of the JPDA method. This amalgamation of sequential track initiation and track termination within the JPDA method allows for adaptive adjustment of the number of targets detected by different sensors. Finally, the efficiency of the proposed method is demonstrated in different simulations.