无人机多源数据在路面裂缝细部测量中的应用研究

打开文本图片集

中图分类号：TP391.41;U418.6 文献标识码：A 文章编号：1672-3791（2026）01-0062-05

Research on Application of Drone Multi-Source Data in Detailed MeasurementofRoad Crack

LEI Gang1 LIU Chaoqun1* XIE Dongdong² HE Yinxin1 LIU Yang1 1.Sichuan Communication Surveying & Design Institute Co.，Ltd.， Chengdu， Sichuan Province， 610017 China; 2.Guizhou Institute of Mountain Resources，Guiyang，Guizhou Province，55ooo1 China

Abstract： With the development ofdrone technology，its application in highwayand ancillry detail surveying has become increasingly widespread.Based on the DJI Warp M3oo drone platform，integrating high-precision photogrammetry and LiDAR equipment，multi-source data acquisitionand mapping of road cracks on a highway is carried out.Multi-dimensional data results such as digital orthophoto map（DOM），digital surface model （DSM），inclined 3D model and clasified point cloud are obtained through fusion processing. Combined with YOLOv5 deep learning algorithm，aprogresive analysis framework of“target detection-3Dmeasurement-trend prediction"is constructed to achieve colaborativeoptimization ofcrack detection accuracy and effciency.The experimental data show that the eficiencyofthis method is more than 15 times higher than thatoftraditional manual detection，with a mean average precision reaches 92.7% .The research results confirm that the combination of multi-source data fusion and deep learning technology can efectively promote the development of road maintenance monitoring in the direction of intelligence.

Keywords： Drone measurement; Multi-source data; Road crack; YOLOv5 deep learning algorithm

公路路面与附属设施的精准测量是保障道路安全与维护效率的关键环节。（剩余5958字）