图像智能识别技术在高速铁路基础设施检测中的应用  

Application of Intelligent Image Recognition Technology for Detection in HSR Infrastructure

作　　者：严鹏[1] 廖峪 陈伟庚 刘晓江 杨长卫[1] YAN Peng;LIAO Yu;CHEN Weigeng;LIU Xiaojiang;YANG Changwei(Southwest Jiaotong University,Chengdu Sichuan 610031,China;Shenzhen Engineering Construction Headquarter,China Railway Guangzhou Group Co Ltd,Shenzhen Guangdong 518000,China;China Railway Erju 4th Engineering Group Co Ltd,Chengdu Sichuan 610306,China)

机构地区：[1]西南交通大学,四川成都610031 [2]中国铁路广州局集团有限公司深圳工程建设指挥部,广东深圳518000 [3]中国中铁二局第四工程有限公司,四川成都610306

出　　处：《中国铁路》2019年第11期109-113,共5页China Railway

基　　金：四川省科技研究开发计划项目（18MZGC0186、18MZGC0247）

摘　　要：基础设施病害问题已成为威胁运营安全的焦点,如何准确、及时发现病害是迫切需要解决的难题。借助人工智能和大数据分析技术,采用目前较为成熟的图像识别算法,以接触网悬挂状态及缺陷为对象,阐述图像识别技术在高速铁路基础设施检测方面的应用。结果表明:图像识别技术在高速铁路基础设施智能化检测方面具有一定的普适性,能够创造大量的虚拟劳动力,克服人工疏忽等主观因素,工作效率有效提升。该研究成果可为人工智能技术在铁路上的应用提供参考。The problem of infrastructure disease has been the focus which threat to operation safety.How to detect the disease accurately and timely is an urgent problem to be solved.With the help of artificial intelligence and big data analysis technology,the application of image recognition technology in the detection of high speed railway(HSR)infrastructure is described with the suspension status and defects of catenary as the object,using the relatively mature image recognition algorithm at present.The results show that the image recognition technology has a certain universality in the intelligent detection of HSR infrastructure which can create a large number of virtual labor force,overcome subjective factors such as artificial negligence,and effectively improve the work.The research results can provide reference for the application of artificial intelligence technology in railway.

关 键 词：高速铁路 图像识别 基础设施 检测 人工智能 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置] U298[自动化与计算机技术—控制科学与工程]