基于机器视觉的钢轨磨耗检测系统研究  

Research on Rail Wear Measurement System Based on Machine Vision

作　　者：唐晓敏 王培俊[1] 吕东旭[1] 李文涛[1] 

机构地区：[1]西南交通大学机械工程学院,四川成都610031

出　　处：《仪表技术与传感器》2017年第9期59-62,87,共5页Instrument Technique and Sensor

基　　金：国家自然科学基金项目(51305368);四川省科技支撑计划项目(2013GZX0154)

摘　　要：为提高钢轨磨耗检测的速度和精度,设计了基于机器视觉的钢轨磨耗检测系统,用于取代人工检测。利用机器视觉和计算机技术,通过硬件平台的搭建和软件的设计实现了钢轨磨耗在线检测。针对道岔区域的匹配问题,提出了先通过轮廓矩识别道岔和钢轨轮廓,再结合特征匹配和SVD-ICP算法完成道岔匹配的新方法。通过实验测试,证明该系统有较高的测量精度和效率,可以较好地满足钢轨磨耗在线检测需要,同时,对于道岔有较好的适应性,能够实现道岔尖轨的匹配,为进一步实现更为复杂的道岔的检测提供了数据支持。To improve the detection precision and speed of the rail wear detection,a machine vision based rail wear measurement system was developed to take the place of human inspector.The system uses machine vision and computer technology to implement the rail wear on line with the foundation of its hardware platform and software.For the matching problem in the turnout area,contour moment was used for distinguishing turnout and rail profile,and then characteristic points and SVD-ICP were combined to implement turnout detection in the paper.Finally,the experiment shows that the system has high accuracy and efficiency and can meet the demand of the online detection of the rail wear need.At the same time,it has better adaptability to the turnout area and can realize the match of the switch railroad,which provides the data support for the further detection of turnout with more complex rail profile.

关 键 词：钢轨磨耗 机器视觉 特征点 最近点迭代 道岔 

分 类 号：U213[交通运输工程—道路与铁道工程]