Offline Estimation and Online Update Algorithm of Clutter Intensity in Multi-target Tracking

Aimed at the problem that the offline estimation algorithm of clutter intensity has lower accuracy in multitarget tracking, we propose an online update algorithm of clutter intensity. In our method, the update of the clutter intensity involves the average clutter number per scan and clutter spatial distribution. During the tracking procedure, the average clutter number per scan is updated via the clutter number. Moreover, the confidence degree method is adopted to divide the clutter set and the parameters of clutter spatial distribution are updated. In addition, the k-means method is further employed to modify the clutter spatial distribution. At last, the smoothing operation is introduced to improve the robustness of the proposed algorithm. Numerical simulations show that our algorithm has higher estimate accuracy of clutter intensity and better performance in multi-target tracking over the existing algorithm.