轴承端面缺陷的视觉检测方法.pdfVIP

学兔兔 !兰 ! 二 2 轴承 2016年3期 60—63 CN41—1148／TH Bearing 2016，No．3 轴承端面缺陷的视觉检测方法 王恒迪 ，李莎 ，杨建玺 ，刘盟盟 (1．河南科技大学 机电工程学院，河南 洛阳 471003；2．中国一拖集团有限公司计量检测中心，河南 洛阳 471003) 摘要：针对轴承端面缺陷的在线检测要求及传统人工检测方法的不足，提出了一种基于机器视觉检测技术的轴 承端面缺陷检测方法，能够通过对轴承端面图像的处理与分析，快速、准确地实现轴承端面缺陷的识别。首先 使用CCD数字摄像机作为图像传感器进行图像采集；然后通过二值化处理、边缘检测处理对图像进行快速定 位与分割；最后采用8连通域标记法进行缺陷识别。试验表明：该系统运行稳定，运算速度快，抗噪能力强，实 时性好，可有效检测出轴承端面缺陷。 关键词：滚动轴承；机器视觉；图像处理；缺陷识别 中图分类号：TH133．33；TP391 文献标志码：B 文章编号：1000—3762(2016)o3—0060—03 Visual Inspection Method for End Face Defects on Bearings Wang Hengdi ，Li Sha ，Yang Jianxi ，Liu Mengmeng (1．School of Mechatronics Engineering，Henan University of Science and Technology，Luoyang 47 1003，China； 2．Calibrating and Measuring Center，YTO Group Corporation，Luoyang 471003，China) Abstract：In view of requirements for on—line inspection of end face defects on bearings and sho~ages of traditional manual inspection method，an inspection method is proposed based on machine vision inspection technology．The end face defects on bearings are recognized quickly and accurately by processing and analyzing images of end face of bear— ings．Firstly，the CCD digital camera is regarded as all image sensor to collecting image．Secondly，the image is loca— ted and segmented rapidly by binarization and edge detection processing．Finally，the defects are recognized by using eight—connected components labeling method．The test shows that the system is characterized with stable operation， fast computational speed，strong anti—noise capability and g