应用于低空防御的实时多无人机检测与跟踪算法  

Real Time Multi-Drone Detection and Tracking for Low-Altitude Defense

作　　者：邓俊文 DENG Junwen(The 3rd Research Institute of China Electronics Technology Group Corporation,Beijing 100015,China)

机构地区：[1]中国电子科技集团公司第三研究所,北京100015

出　　处：《电视技术》2023年第6期20-23,共4页Video Engineering

摘　　要：针对低空环境下慢速飞行小型无人机检测与跟踪困难的问题,设计并实现一种基于深度学习的计算机视觉算法,实现对低慢小无人机的快速检测和稳定跟踪。采用基于深度学习的YOLOv8算法快速检测图像中的无人机目标,再用SORT多目标跟踪算法对前后帧的检测结果进行关联,生成目标航迹,实现稳定跟踪。所提算法运行在NVIDIA Jetson Orin硬件平台上。实验结果表明,该算法能实时检测并稳定跟踪目标,对防御低空慢速飞行小型无人机具有较高应用价值。In response to the difficulty of detecting and tracking slow flying small unmanned aerial vehicles in low altitude environments,this paper designs and implements a deep learning based computer vision algorithm to achieve fast detection and stable tracking of low speed small unmanned aerial vehicles.The article uses the YOLOv8 algorithm based on deep learning to quickly detect drone targets in the image,and then uses the SORT multi target tracking algorithm to correlate the detection results before and after frames,generate target tracks,and achieve stable tracking.The proposed algorithm runs on the NVIDIA Jetson Orin hardware platform.The experimental results show that this algorithm can detect and stably track targets in real-time,and has high application value for defending low altitude slow flying small unmanned aerial vehicles.

关 键 词：深度学习 无人机 神经网络 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]