H-PMHT Based Track-Before-Detect Algorithms For Maritime Dim Targets

With the miniaturization and invisibility of maritime targets,the track-beforedetect algorithm for maritime dim targets attracts more and more attention in civilian and military fields.In actual maritime scenarios,the detection and tracking of dim targets is usually faced with problems such as the low signal-to-noise ratio of the dim target,the sensor motion with location uncertainty,and the maneuverable movement of non-cooperative dim targets.In view of the above problems,this manuscript studies the detection and tracking of maritime dim targets based on the histogram probabilistic multi-hypothesis tracking algorithm,and the main works are as follows:1.Aiming at the detection and tracking of maritime dim targets with precisely known sensor location,the track-before-detect algorithm based on the improved histogram probabilistic multi-hypothesis tracking is proposed.The low threshold constant false alarm rate detection and K-means clustering are invoked to the new trajectories initialization in the traditional histogram probabilistic multi-hypothesis tracking framework,improving the quality of the initial target trajectories.In the simulation,the proposed algorithm is compared with the track-before-detect algorithm based on the traditional histogram probabilistic multi-hypothesis tracking.The result shows that the proposed algorithm obtains the more accurate initial target trajectories,hence the target confirmation delay is reduced.2.To solve the problem of maritime dim targets detection and tracking with sensor location uncertainty,the track-before-detect algorithm based on the state augmentation and the histogram probabilistic multi-hypothesis tracking is proposed.The target and sensor states are simultaneously estimated in the iteration.The estimated sensor state is used to update the coordinates of the resolution cells in the measured amplitude intensity image in each iteration,hence the more accurate synthetic point measurements and synthetic measurement covariances are obtained in the next iteration.The simulation and real data verify that the proposed algorithm can reduce the sensor location error and accurately estimate the target states.3.In order to solve the problem of maneuvering maritime dim target detection and tracking with sensor location uncertainty,the track-before-detect algorithm based on the multiple model set-membership filter and the histogram probabilistic multihypothesis tracking is proposed.Since the non-cooperative maneuvering target motion model is time-varying and unknown,the target states are parallelly estimated under multiple motion models and limited to the reasonable range via the multiple model setmembership filter.The final estimated target states are obtained according to the weighted interaction.The simulation shows that the proposed algorithm can effectively detect and output the maneuvering dim target trajectories with sensor location uncertainty.