摘要
针对目前国内轴承钢球直径数字化分选工作效率低、精度差的难题,提出了一种基于机器视觉的钢球直径测量方法。通过双远心镜头系统获取钢球灰度轮廓,对钢球灰度轮廓进行极坐标转换以获取特征点的灰度值,基于多项式拟合灰度曲线并求取亚像素边缘点的位置;对钢球直径进行参数化处理,得到数字化的钢球直径值并提出一种去钢球表面垃圾的优化算法以消除垃圾对测量结果的影响;对视觉系统进行标定,消除镜头畸变产生的测量误差;实际测量结果表明,该算法的重复测量精度小于0.5μm,适用于大批量的在线测量,满足工业自动化测量的需求。
To address the current challenges of low efficiency and poor accuracy in digital sorting of bearing steel ball diameter in China,a method for measuring steel ball diameter is proposed based on machine vision.The grayscale contours of steel balls are obtained through a dual-telecentric lens system.The polar coordinate transformation of grayscale contours of steel balls is performed to obtain the grayscale values of feature points.The grayscale curve is fitted based on polynomials,and the positions of sub-pixel edge points are acquired.The steel ball diameter is parameterized to derive a digital value,and an optimized algorithm for removing contaminants from steel ball surfaces is proposed to eliminate the influence of contaminants on measurement results.The vision system is calibrated to eliminate the measurement errors induced by lens distortion.The actual measurement results demonstrate that the repeatability measurement accuracy of this algorithm is less than 0.5μm,which is suitable for high-volume online measurement and meets the requirements of industrial automation measurement.
作者
李帅奇
陈於学
LI Shuaiqi;CHEN Yuxue(School of Mechanical Science&Engineering,Huazhong University of Science and Technology,Wuhan 430074,China)
出处
《轴承》
北大核心
2025年第6期60-66,共7页
Bearing
关键词
滚动轴承
机器视觉
钢球
分选
标定
在线测量
rolling bearing
machine vision
steel ball
sorting
calibration
on-line measurement
