Multisensor Fusion Algorithms for Tracking

In this paper we extend a multitarget tracking algorithm for use in multisensor tracking situations. The algorithm we consider is Joint Probabilistic Data Association (JPDA). JPDA is extended to handle an arbitrary number of sensors under the assumption that the sensor measurement errors are independent across sensors. We also show how filtering can be handled in multisensor JPDA (MSJPDA) without leading to an exponential increase in filtering complexity. Simulation results are presented comparing the performance of the MSJPDA with another multisensor fusion algorithm and with the single-sensor JPDA algorithm.