面向无人机航拍目标检测的YOL0系列算法研究进展

打开文本图片集

中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2025）20-0027-08

Abstract：UAVs integrated with ObjectDetectiontechnologyleverage theiraerial perspectiveand highmobilityto provide significantfexibilityforjectlocalizationanddatacollction.However，inpracticalaplications，UAVaerialphotoraphy based Object Detection faces challnges such as stringentreal-time requirements，ahigh proportionofsmallobjects，and compleximagebackgrounds，limiting its development and application.Therefore，designing algorithm models capableof efectivelyadapting tothecharacteristicsofUAVaerialimageryis particularlyimportant.Sinceitsiception，theYOLOseries algorithmshaveundergonemultipleterations.Itachevessimultaneousobjectclassfcationandlocalizationviaasinglenetwork forwardpass，ofering excellentreal-time performane.This papersummarizestheresearchprogressof themain versionsof the YOLO seriesalgorithms and presents theirapplications in UAVaerial photography-based ObjectDetection，aiming toprovidea reference for subsequent research.

Keywords： UAV aerial photography; Object Detection; YOLO; Computer Vision; smal Object Detection

0 引言

近年来，随着无人机技术的飞速发展和计算机视觉技术的进步，无人机在目标检测领域展现出巨大应用潜力，无人机航拍目标检测逐渐在智慧城市建设、灾害监测和救援等场景发挥重要作用。（剩余15921字）