基于人工智能的侦测无人机算法研究综述  

Overview of Detecting UAV Algorithms Based on AI

作　　者：王宇龙 双睦融 周丽莎 左顺 何新 曹瑞臣 WANG Yulong;SHUANG Murong;ZHOU Lisha;ZUO Shun;HE Xin;CAO Ruichen(Information Center of China North Industries Group Corporation,Beijing 100089,China;State Key Laboratory of Dynamic Measurement Technology,North University of China,Taiyuan 030000,China)

机构地区：[1]中国兵器工业信息中心,北京100089 [2]中北大学省部共建动态测试技术国家重点实验室,山西太原030000

出　　处：《测控技术》2024年第9期35-44,共10页Measurement & Control Technology

基　　金：国家自然科学基金(62075199);山西省重点研发计划项目(201803D121050)。

摘　　要：随着科技的进步与生产生活质量的稳步提升,无人机在工业领域的重要性逐步显露,与此同时,利用无人机进行非法活动的数量也逐年上升,对无人机进行侦测与识别迫在眉睫。传统的对无人机进行侦测的算法以目标匹配为基础,需要建立庞大的目标库进行复杂的匹配计算,存在目标误报率过高、漏报率过高,识别时间过长的缺陷。近年来,深度学习技术在图像处理、语音识别、自然语言处理等领域取得了突破性进展,在此背景下,利用深度学习技术的特征学习能力能够自动学习目标特征,提升目标识别的精准性,从而为对无人机的侦测和识别提供了一种新的途径。通过搜集该领域最新研究成果,从雷达、声学信号、视觉信号、射频信号等技术角度出发,结合智能技术,对当前识别算法进行了归纳与阐述,最后对该领域中的研究所面临的主要挑战进行了总结,并展望了未来的研究重点。With the progress of science and technology and the steady improvement of the quality of production and life,the importance of unmanned aerial vehicle(UAV)in the industrial field is gradually revealed.At the same time,the number of illegal activities using UAV is also increasing year by year,and the detection and i-dentification of UAVs is imminent.The traditional detecting UAV algorithm is based on target matching,which needs to build a huge target library for complex matching calculation,and faces the defects of high target false alarm rate,high false negative rate and long recognition time.In recent years,deep learning technology makes breakthroughs in image processing,speech recognition,natural language processing and other fields.In this context,using the feature learning ability of deep learning technology can automatically learn target features to improve the accuracy of target recognition,which provide a new way for UAVs'detection and recognition.By collecting the latest research results in this field,the current recognition algorithms is summarized and expoun-ded from the technical point of view of radar,acoustic signal,visual signal and radio frequency,combined with intelligent technology.Finally,the main challenges facing the research in this field are summarized,and the fu-ture research priorities are prospected.

关 键 词：侦测无人机 识别 人工智能 深度学习 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]