Abstract:

For multi target tracking in clutter, the probability hypothesis density (PHD) filter can not provide the track information of each target, thus, a novel method based on PHD filter and data association is proposed. Firstly, the PHD filter is utilized to remove the clutters and get the estimated target number and states at each time step. Then, the results of PHD filter are redefined as the measurements, which are managed by data association so that the false alarms and miss detections are further eliminated and the tracks of targets are provided. Simulation results demonstrate that the proposed method can improve the accuracy of multitarget tracking effectively as well as provide track information of each target.