基于机器视觉的大轴端面锈蚀检测与清洗  

Rust Detection and Cleaning of Large Shaft End Face Based on Machine Vision

作　　者：张宏发 刘复兴 周诗洋 杨金堂[1,2] Zhang Hongfa;Liu Fuxing;Zhou Shiyang;Yang Jintang(Key Laboratory of Metallurgical Equipment and Control,Ministry of Education,Wuhan University of Science and Technology,Wuhan 430081,China;Hubei Key Laboratory of Mechanical Transmission and Manufacturing Engineering,Wuhan University of Science and Technology,Wuhan 430081,China;Steelmaking Plant,E’cheng Iron and Steel Co.,Ltd.,Baowu Group,E'zhou 436099,China)

机构地区：[1]武汉科技大学冶金装备及其控制教育部重点实验室,武汉430081 [2]武汉科技大学机械传动与制造工程湖北省重点实验室,武汉430081 [3]宝武集团鄂城钢铁有限公司炼钢厂,湖北鄂州436099

出　　处：《煤矿机械》2021年第10期170-172,共3页Coal Mine Machinery

基　　金：国家重点专项资助项目(2018YFC1902400);国家自然科学基金(51805386)。

摘　　要：为解决传统清洗方式所带来的高成本、低效率、污染严重等问题,提出一种基于机器视觉的激光清洗视觉检测方法。通过灰度化、直方图均衡和中值滤波对图像进行预处理,采用灰度级形态学重建消除阴影和反射光,使用最大类间方差法及阈值分割对图像进行特征提取,使用自拟特征判断函数来完成对清洗效果的最终识别。实验结果表明,该方法能准确检测出大轴端面清洗率,单次检测时间控制在0.17 s内,满足其精度及效率要求。In order to solve the problems of high cost, low efficiency and serious pollution caused by traditional cleaning methods, a laser cleaning vision detection method based on machine vision was proposed. Used the gray level transformation, histogram equalization and median filter to achieve the preprocessing. Used the grayscale morphology reconstruction algorithm to eliminate the shadow and the reflected light. Used OTSU algorithm and threshold segmentation algorithm to extract image feature.Used a self-designed feature judgment function to recognize the cleaning effect. The experimental results show that the method can accurately detect the cleaning rate of the large shaft end face, and the single detection time is controlled within 0.17 s, which meets the requirements of accuracy and efficiency.

关 键 词：锈蚀检测 机器视觉 最大类间方差法 形态学重建 

分 类 号：TH133.2[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]