基于机器视觉的铁路混凝土桥梁裂缝检测系统研究与应用  

Research and Application of Crack Detection System for Railway Concrete Bridge Based on Machine Vision

作　　者：魏剑峰 李艳 禚一 邸昊 WEI Jianfeng;LI Yan;ZHUO Yi;DI Hao(China Railway Design Corporation,Tianjin 300308,China)

机构地区：[1]中国铁路设计集团有限公司,天津300308

出　　处：《铁道建筑技术》2023年第9期20-23,31,共5页Railway Construction Technology

基　　金：中国国家铁路集团有限公司科技研究开发计划项目(K2021G013);中国铁路设计集团有限公司科技开发课题(2022B03406021,2022A02480005,2021B240630)。

摘　　要：铁路混凝土桥梁是我国高速铁路建设的主力梁型,建设规模巨大。在铁路混凝土桥梁的建造和运维过程中,混凝土桥梁的裂缝检测是一项十分重要的工作。为了降低检测人员的劳动强度,解决裂缝检测效率低、精度差的问题,本文对基于机器视觉技术的铁路混凝土桥梁的裂缝检测系统开展研究与应用。搭载着高分辨率相机的图像采集子系统用于桥梁裂缝图像采集,提出了一种基于机器视觉技术的裂缝智能识别算法,算法以裂缝检测模块的形式集成于系统当中,能够对采集的桥梁图像中的裂缝进行实时检测。系统在铁路梁场简支梁破坏试验中进行了应用,检测精度和时间均能满足试验规范的要求,应用效果良好。Railway concrete bridge is the main beam type of high-speed railway construction in our country,and the construction scale is huge.In the process of construction and operation of railway concrete bridge,the crack detection of concrete bridge is a very important work.In order to reduce the labor intensity of inspection personnel and solve the problems of low efficiency and poor accuracy of crack detection,this paper researches and applies the crack detection system of railway concrete bridge based on machine vision technology.The image acquisition subsystem equipped with high-resolution camera is used for bridge crack image acquisition.In this paper,an intelligent crack recognition algorithm based on machine vision technology is proposed.The algorithm is integrated into the system in the form of crack detection module,which can detect cracks in the collected bridge images in real time.The system has been applied in the failure test of simple supported beams in railway beam yard.The test accuracy and time can meet the requirements of test specifications,and the application effect is good.

关 键 词：铁路桥梁 机器视觉 深度学习 裂缝检测 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] U446[自动化与计算机技术—计算机科学与技术]