转向架轮对参数的结构光视觉测量方法与实现  

Structured Light Vision Measurement and Realization of the Wheel Set Parameters

作　　者：洪磊 王保升[2] 杨小兰[3] HONG Lei;WANG Baosheng;YANG Xiaolan(School of Automotive and Rail Transit,Nanjing Institute of Technology,211167,Nanjing,China)

机构地区：[1]南京工程学院汽车与轨道交通学院,南京讲师211167 [2]南京工程学院智能制造装备研究院,南京211167 [3]南京工程学院机械工程学院,南京211167

出　　处：《城市轨道交通研究》2018年第7期129-133,共5页Urban Mass Transit

基　　金：国家自然科学基金(61703200);江苏省产学研合作前瞻性联合研究项目(BY2016008-04);南京工程学院科研基金项目(PTKJ201606);江苏省"青蓝工程"资助项目

摘　　要：为实现转向架轮对半径参数的精确检测,提出了一种基于结构光视觉传感器的非接触式检测方法。该方法首先采集轮对的线结构光条纹图像,采用中值滤波和最大间类方差法对图像进行去噪和阈值分割后,运用细化及Hough变换算法提取结构光条纹中心直线,最后结合视觉传感器标定参数计算出轮对特征点坐标及尺寸参数。通过轮对轮缘半径检测试验,验证了该方法的误差小于0.05mm,具有较好的检测精度,满足实际现场检测的要求。In order to realize the accurate measurement of wheel set radius parameters, a non-contact detection method based on structured light vision sensor is proposed. In this method, the stripe image of wheel set is firstly acquired, then through denoising and threshold segmentation of the stripe image by using median filtering and Otus method （maximum between cluster variance）, the center of structured light stripe line is extracted based on thinning and Hough transform algorithm. Finally, according to the vision sensor calibration parameters, the coordinate and dimension parameters of the characteristic points are calculated. A wheel rim radius detection experiment shows that the error in the method is less than 0.05 mm, it proves to have favorable detection precision and could meet actual requirements for field test.

关 键 词：转向架轮对 结构光视觉 测量方法 非接触式 

分 类 号：U260.331.1[机械工程—车辆工程]