Radar Weak Target Tracking Based on LMB Smoothing

Abstract

The effective detection and tracking of low, small and slow targets is a current hot issue in the radar field. The paper is a sequel to radar weak target detection based on RFT and AMF fusion focus, and concentrates on the tracking of typical UAVs by pulse Doppler radar. The implementations of tracking algorithm mainly include: first, the preprocessing of the plots centroid processing and the multi-frame acceleration estimation has been applied. Then the labeled multi-Bernoulli (LMB) smoother is used for the multi-target tracking. Finally, the final trajectories of targets are output after trajectory continuity judgment and effective trajectory extraction. The LMB smoother can output the smoothed trajectories of targets, therefore, improving the estimation accuracy of distance and velocity. Trajectory continuity judgment and effective track extraction can effectively overcome the problems of breaking trajectory caused by velocity jitter and false trajectory caused by velocity ambiguity. The comprehensive score rate to process multiple test data is 96.67%, which verifies the effectiveness and robustness of the proposed tracking algorithm.

Keywords

• |pulse doppler radar|trajectory continuity judgment|lmb|filtering smooth|target tracking