火力与指挥控制2025，Vol.50Issue(12)：150-158,9.DOI:10.3969/j.issn.1002-0640.2025.12.020
基于多机协同的无人机目标识别方法
Target Recognition Methods for UAVs Based on Multi-UAV Cooperation
摘要
Abstract
Addressing the military needs of unmanned aerial vehicles(UAVs),a multi-UAV cooperation target detection system is designed.This system achieves distributed bombing operations among multiple UAVs by actively acquiring location information through route planning for patrol UAVs.By introducing deep deformable convolutions and incorporating a small target feature aggregation network,the system enhances the ability to express features of small targets.The Ghost Module is introduced to implement a lightweight design for the model YOLOv8,enabling precise identification of combat areas.The experiments have demonstrated that the improved YOLOv8 algorithm,under multi-UAV cooperation,achieves an average recall rate and global accuracy of 90.1%and 95.6%,respectively,with a computation speed of 24 frames per second.关键词
多机协同/目标探测/YOLOv8算法/轻量化设计/对地攻击/无人机Key words
multi-UAV collaboration/target detection/algorithm YOLOv8/lightweight design/ground attack/UAVs分类
军事科技引用本文复制引用
SHI Changhai,DONG Tiantian,LUO Xin,WANG Bangzhao,XIA Yuguo..基于多机协同的无人机目标识别方法[J].火力与指挥控制,2025,50(12):150-158,9.基金项目
江苏省自然科学基金(24KJD510004) （24KJD510004）
江苏省高等学校基础科学(自然科学)研究重大项目(124KJA510005) （自然科学）
江苏信息职业技术学院校级重点科研团队技术开发项目 （）
江苏高校"青蓝工程"基金资助项目 （）
