Graph-based Track Stitching Method for Low Altitude Marine Multi-target Tracking

In the low altitude marine multi-target tracking scenario, the multiple hypothesis tracking (MHT) algorithm generates a large number of track segments due to the influence of sea clutter. To address this problem, an amplitude information aided graph-based track stitching method is proposed. First, the method models the track stitching scenario as a graph; after that, the amplitude information of different targets are passed in the track graph to obtain the association likelihood of amplitude information between track segments. Under the assumption of Markov, the association likelihood of the amplitude information between the track segments is multiplied with the association likelihood of the target states to obtain the association likelihood between the track segments. Finally, the minimum-cost maximum-flow (MCMF) algorithm is used to solve the stitching results. The simulation results show that the proposed algorithm can effectively improve the track stitching performance in the low altitude marine multi-target tracking scenario, and has a certain degree of improvement compared with the Hungarian algorithm in terms of rates of false association and target fragmentation.