现代雷达2025，Vol.47Issue(11)：1-14,14.DOI:10.16592/j.cnki.1004-7859.2025062001
面向雷达无人机目标检测识别的深度学习方法综述
A Survey of Deep Learning Methods for Radar UAV Target Detection and Recognition
摘要
Abstract
With the widespread application of unmanned aerial vehicle(UAV),a series of increasingly prominent public safety and privacy issues have emerged while promoting efficiency improvement and innovative applications in multiple fields.Therefore,there is an urgent need to regulate UAV flight activities.Currently,machine learning methods have become mainstream ones for detec-tion and recognition of UAV targets.In this paper,deep learning methods for radar-based detection and recognition of UAV targets are focused.Firstly,the three major feature extractors in deep learning—convolutional neural networks,recurrent neural networks,and Transformer networks,are reviewed.Then,a systematic review on the applications and variants of these networks in radar-based detection and recognition of UAV targets is provided,covering the design features and typical performance characteristics of different structures.Finally,future trends and challenges in radar-based detection and recognition technologies for UAV targets are discussed.In this paper,the relevant deep learning methods are introduced,while the corresponding datasets and typical experi-mental setups are also described,which is greatly benefical for comprehensively and systematically understanding the application of deep learning methods in radar-based detection and recognition of UAV targets,and helps to promote the further development of ra-dar-based detection and recognition of aerospace targets.关键词
无人机/雷达目标/目标检测/目标识别/深度学习Key words
unmanned aerial vehicle(UAV)/radar target/target detection/target recognition/deep learning分类
信息技术与安全科学引用本文复制引用
周玉军,欧阳可赛,程强..面向雷达无人机目标检测识别的深度学习方法综述[J].现代雷达,2025,47(11):1-14,14.基金项目
国家自然科学基金企业创新发展联合基金重点资助项目(U22B2059) （U22B2059）
