基于AI的路面病害自动识别系统研发与应用  

Research and Application of Automatic Identification System for Pavement Defects Based on AI

作　　者：丁攀 翟晓成 苟晓锋 刘俊杰 刘士宽 费昀 DING Pan;ZHAI Xiao-cheng;GOU Xiao-feng;LIU Jun-jie;LIU Shi-kuan;FEI Yun(Lanzhou Highway Development Center of Gansu Province,Lanzhou 730000,China;CHINA TRANS GEOMATICS Co.Ltd.,Beijing 100094,China)

机构地区：[1]甘肃省兰州公路事业发展中心,兰州市730000 [2]中交宇科(北京)空间信息技术有限公司,北京市100094

出　　处：《公路》2024年第9期336-342,共7页Highway

基　　金：2022年甘肃省交通厅科研课题《公路路面巡查智能识别技术应用研究》,项目编号GSXYX-ZNXJ-202301。

摘　　要：在日常养护路面病害识别方面,大多采用肉眼观测实际路面的方式,随着科技的发展,为降低劳动强度,发展为利用路面影像人工判读进行病害识别。无论怎样,人工用肉眼识别病害都存在漏检现象,准确率低,工作效率低。为解决此问题,以AI代替肉眼进行路面病害识别,并进行路面病害的可视化管理。基于AI的路面病害自动识别系统研发与应用,显著提升了路面病害识别效率,为路面病害自动化识别技术的发展做出了贡献,提高了养护管理科技水平。In terms of daily inspection of road surface,the mode of manually filling in inspection forms is frequently adopted.With the development of technology,road images are used for manual interpretation of defects identification in order to reduce labor intensity.However,the human recognition method may cause missed detection,low accuracy,and low work efficiency.Based on this,AI is used to replace human eyes for recognition of road surface defects,and provides visual management of pavement diseases.The research and development and application of Al-based automatic recognition system for road defects have significantly improved the efficiency of road defects recognition,contributed to the development of automatic recognition technology for road diseases,and improved the level of maintenance management technology.

关 键 词：日常养护巡查 人工智能 路面病害自动识别 病害可视化 

分 类 号：U416.06[交通运输工程—道路与铁道工程] U495