Decoupling joint probabilistic data association algorithm for multiple target tracking

Abstract

The authors presents a decoupling joint probabilistic data association (DJPDA) algorithm by decoupling the JPDA algorithm into separate PDA algorithms. To demonstrate the decoupling idea, they derive a set of complete decoupling PDA equations for two and three targets. Since these equations are impractical for real-time systems and are unavailable for an arbitrary number of targets, an approximate algorithm is given. The proposed decoupling algorithm is evaluated using computer simulation. It is shown that the performance of the DJPDA algorithm is close to that of the JPDA algorithm but it needs much less computation.