Application of a two-scan data association algorithm to multiple target tracking

In view of the computational burden of a traditional joint probabilistic data association algorithm, a new two-scan data association method based on an interior point linear programming approach was introduced. The principle of the new algorithm was analyzed, and its performance was compared with one of the traditional joint probabilistic data association algorithm. It can be seen that the new algorithm is more effective from the aspects of theory. The new method can be applied to multiple target tracking. The improved accuracy and speed of the new method proved them to be superior to those of the traditional by computer simulation experiments.