Joint probabilistic data association methods avoiding track coalescence

Abstract

For the problem of tracking multiple targets the joint probabilistic data association (JPDA) filter has shown to be very effective in handling clutter and missed detections. The JPDA, however, also tends to coalesce neighbouring tracks. Through comparing JPDA with the exact nearest neighbour PDA (ENNPDA) filter, Fitzgerald has shown that hypotheses pruning is an effective way to prevent track coalescence. The dramatic pruning used for ENNPDA however leads to an undesired sensitivity to clutter and missed detections. In this paper new algorithms are derived which combine the advantages of JPDA and ENNPDA. The effectiveness of the new algorithms is shown through Monte Carlo simulations.