A new track-before-detect method based on plot and track filtering with multilevel features

Aim at weak radar target detection and tracking in complicated background, a track-before-detect (TBD) method based on plot and track filtering is proposed. First, the plots obtained from low threshold processing are tracked and filtered, then, based on the plot and track features, multi-level thresholds are designed to eliminate false plots and tracks, finally, the target detection and tracking results can be obtained simultaneously. Compared with the conventional TBD algorithm, the proposed algorithm is robust and with less computation cost, thus, it is of good value for engineering application. Finally, the validity of the algorithm is verified by real data experiment.