光学精密工程2025，Vol.33Issue(9)：1446-1455,10.DOI:10.37188/OPE.20253309.1446
基于知识图谱的红外目标部件识别
Knowledge graph-based method for infrared target component recognition
摘要
Abstract
To address challenges in component recognition,such as self-occlusion,unclear visual fea-tures,and substantial feature variation due to distance,a novel infrared target component recognition meth-od based on a knowledge graph is proposed.This method utilizes a whole-to-component prediction(WCP)strategy to sequentially recognize target components.Initially,the overall target is detected,fol-lowed by an expansion of the target region to high resolution to enhance signal details.Subsequently,a component-related attention module(CAM)integrated with the knowledge graph exploits structural rela-tionships among parts to infer visible interconnections and employs attention mechanisms to improve recog-nition performance,thereby mitigating issues arising from ambiguous visual features.For components af-fected by self-occlusion,a Self-Occlusion Learning Rate Decay(SLD)control strategy,based on self-re-moval capability,is introduced to strengthen the model's capacity to learn from occlusions and facilitate convergence.Validation is conducted using an indoor target equivalence scaling system,employing room-based models across various orientations and distances,with aircraft tested under diverse conditions,achieving an average precision of 92.2%.Experimental results demonstrate that the proposed method sur-passes existing approaches in component recognition accuracy and recall,markedly enhancing both preci-sion and recall metrics.关键词
红外目标识别/知识图谱/整体识别/全局向量模型/注意力模块/学习率控制Key words
infrared target recognition/knowledge graph/whole to component predict/attention mod-ule/self-occlusion learning rate decay分类
信息技术与安全科学引用本文复制引用
刘海毅,李正周,李傲燃,刘海涛..基于知识图谱的红外目标部件识别[J].光学精密工程,2025,33(9):1446-1455,10.基金项目
国家自然科学基金资助项目(No.61675036) （No.61675036）
十三五国防预研基金资助项目(No.6140415020312) （No.6140415020312）
