面向轴承生产线的视觉检测光源系统  

Visual Detection Light Source System for Bearing Production Line

作　　者：黄振宇 金京 钱淼 HUANG Zhenyu;JIN Jing;QIAN Miao(Faculty of Mechanical Engineering and Automation,Zhejiang Sci-Tech University,Hangzhou 310018,China;Zhejiang Yingfeng Technology Co.,Ltd.,Shaoxing 312073,China)

机构地区：[1]浙江理工大学机械工程学院,杭州310018 [2]浙江迎丰科技股份有限公司,浙江绍兴312073

出　　处：《轴承》2025年第1期88-94,共7页Bearing

基　　金：浙江省“尖兵”“领雁”研发攻关计划资助项目(2023C01158)。

摘　　要：为降低工业自动化生产过程中轴承端面块状瑕疵的漏检率和误检率,对视觉检测常用的光源类型进行了对比试验,在工业环境下采用不同光源获取图像并使用不同的图像增强算法进行优化处理,通过像素点检测的方式检测轴承端面块状瑕疵,以工厂的检测标准进行应用验证。结果表明:同轴光源作为视觉光源时所采集图像在各种图像增强算法处理下的检测效果均好于其他光源;自适应直方图均衡化对轴承端面瑕疵的总体检测效果较好,Gamma变换则对降低漏检率有较好作用。In order to reduce the missed and false detection rates of block defects on bearing end faces during industrial automation production process,a comparative test is conducted on types of light sources commonly used in visual detection.Different light sources are used to capture images under industrial environment,and various image enhancement algorithms are applied for optimization processing.The pixel detection method is utilized to detect block defects on bearing end faces,and the application is verified with factory's detection standards.The results indicate that when the coaxial light source is used as visual light source,the image acquired by coaxial light source exhibits superior detection effect compared with other light sources under different image enhancement algorithms;the adaptive histogram equalization demonstrates excellent effectiveness in overall detection of defects on bearing end faces,while Gamma transformation effectively reduces the missed detection rate.

关 键 词：滚动轴承 机器视觉 光源 表面缺陷 检测 图像处理 

分 类 号：TH133.33[机械工程—机械制造及自动化] TP391.4[自动化与计算机技术—计算机应用技术]