基于航迹特征原型选择的军民船双分支融合识别OA

The accurate classification and recognition of maritime military and civilian targets by radar plays a foundational role in generating situational awareness of sea surfaces and holds significant importance for maritime secu-rity.At present,most of the research is based on trajectory features to identify military and civilian vessels,while the existing methods rely on single-frame static features,which is difficult to deal with the camouflage behavior of military vessels.Or employing single deep temporal models that are susceptible to noise interference.To address these challeng-es,this paper proposes a dual-branch fusion recognition framework for military and civilian vessels based on trajectory feature prototype selection.The method first employs a feature data prototype selection strategy to mitigate category im-balance problems.Then,two parallel branches are constructed:a static branch for instantaneous feature processing and a spatio-temporal branch for temporal pattern analysis.Notably,the spatio-temporal branch incorporates an adaptive di-lated convolutional temporal convolutional network that dynamically adjusts its receptive field according to local trajec-tory smoothness,enabling efficient capture of tactical maneuver patterns.Finally,an adaptive confidence weighting mechanism combined with Dempster-Shafer evidence theory is implemented to achieve context-aware evidence fusion.Experimental results on the real trajectory dataset demonstrate that the proposed approach achieves the accuracy of 95.2%in the identification of military vessels.