KL distance based adaptive particle filter for maneuver target Track-before-detect

An adaptive particle filter Track-before-detect(PF-TBD) algorithm based on Kullback-Leibler distance for Over-the-horizon radar (OTHR) sea surface dim slow maneuvering target detection and tracking is proposed in this paper. First, echo signal is passed through MTI filter for suppress sea clutter. Then, considering maneuver target can broaden doppler spectrum in each coherent integration time (CIT), subsection of echo signal is integrated and doppler spectrum is obtained after windowing processing. Finally, considering the change of state can influence the number of particles in each doppler bin, KL distance based adpative sampling particles in each doppler bin is proposed and achieves a fast detection and tracking algorithm to solve OTHR sea surface dim slow maneuvering target detection and tracking problem. Simulation results show that proposed algorithm can quickly detect and track target without of lost accuracy of target detection and tracking.