基于机器视觉的轴承内外圈尺寸检测及分类  

Detecting and Classifying the Inner-outer Ring Sizes of Bearings Based on Machine Vision

作　　者：范帅[1,2] 汤绮婷[1] 卢满怀[1] 

机构地区：[1]电子科技大学中山学院,广东中山528400 [2]电子科技大学机械电子工程学院,四川成都611731

出　　处：《自动化仪表》2016年第11期77-80,87,共5页Process Automation Instrumentation

摘　　要：在实际生产线上,根据不同的尺寸偏差对同一型号的轴承内外圈进行分类组装。研究了机器视觉技术,设计了一套轴承内外圈尺寸检测和分类系统,以替代人工综检分类。通过设计检测机构,对图像进行滤波、二值化,相机与镜头的标定,形态学去除内壁和边缘提取等方式,达到检测和分类的目的。在尺寸测量部分,通过对两个特殊像素点的运算得到了轴承圆的内外半径,避开了对圆心的定位和轮廓的提取。试验表明,该系统很好地实现了轴承内外圈尺寸的检测和分类。In practical production line of bearings, the same model bearings are classified and assembled according to their different sizes of inner and outer rings. The technology of machine vision is researched, and a set of detection and classification system for inner and outer ring sizes of bearings is designed to replace the manual inspection and classification. Through designing detection mechanism, image binarization,camera and lens calibration; morphology removing the lining,and edge extraction,etc. ,the goals of detection and classification are achieved. In the measurement of sizes,by operating two special pixel points,the inner - outer radius of the bearing is obtained,which avoids the positioning for the center of circle and contour extraction. Test shows that the system well implements the detection and classification of inner and outer ring sizes of bearings.

关 键 词：轴承 机器视觉 图像处理 尺寸检测 相机标定 滤波 二值化 边缘检测 轮廓提取 

分 类 号：TH13[机械工程—机械制造及自动化] TP391[自动化与计算机技术—计算机应用技术]