Track Segment Association Algorithm based on Multi-feature Inference Synthesis

Ensuring the consistency of target batch numbers in adjacent radar detection areas is crucial for multi-target management in complex environments. In this paper, we propose a novel track segment association (TSA) algorithm based on muiti-feature inference synthesis to achieve accurate batch number unification of tracks. Firstly, through a hierarchical multi-feature fusion based on the targets' current and historical signal attribute information, the target type recognition for different track segments is achieved. In particular, considering the lack of comprehensive analysis of multi-feature in traditional TSA algorithms, we integrate the type, motion mode, and kinematic information of targets to realize more accurate and adaptable TSA. Finally, the proposed algorithm is applied to two simulation scenarios and the experimental results indicate multi-feature inference synthesis is efficient in TSA.