高铁白车身装配质量的机器视觉检测方法  

Machine vision inspection method for high-speed rail body-in-white assembly quality

作　　者：侯天龙 孙永奎[1] 马磊[1] 孙小通 HOU Tianlong;SUN Yongkui;MA Lei;SUN Xiaotong(School of Electrical Engineering,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学电气工程学院,四川成都610031

出　　处：《传感器与微系统》2021年第7期142-144,共3页Transducer and Microsystem Technologies

基　　金：2017年度国家重点研发计划“智能机器人”重点专项项目(2017 YFB1303402,2017YFB 1303402-03)。

摘　　要：针对高铁白车身关键零部件装配的间隙宽度、深度和高度差质量检测,提出了一种基于机器视觉的无接触式装配质量检测方法。设计了自适应窗口的中值滤波算法,采用脊线跟踪和Hession矩阵相结合的算法实现了亚像素级的中心线条纹的提取。通过像素坐标系与物理坐标系的映射关系完成对装配特征的尺寸检测。实验结果表明:所提方法检测精度达到了0.05 mm。Aiming at quality inspection measuring gap width,depth and height-difference of key assembly parts of high-speed rail carriage,a non-contact assembly quality detection method based on machine vision is proposed.Firstly,a adaptive window median filter is designed for suppressing the salt-pepper noise.Secondly,a subpixel centralline stripe is extracted using the algorithm of ridge line tracking and Hession matrix.Then,the assembly feature is measured by the mapping relationship between the pixel coordinate system and the physical coordinate system.The experimental results show that the detection accuracy of the method reaches 0.05 mm.

关 键 词：三维测量 结构光 机器视觉 装配质量 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]