对于微小轴承的内径测量,基于机器视觉的非接触测量技术具有广阔的应用前景。针对轴承的边缘像素点,投影将感光像元划分成感光值不同的两个部分,像素单元的最终灰度值为投影两侧局部灰度统计值的面积加权平均值,取该像素的矩形邻域,并对其邻域灰度值进行高斯加权处理,将边缘投影按泰勒公式展开为二次曲线得到高精度亚像素边缘。针对微小轴承进行实验,将本文算法与基于直线拟合的方法进行比较,每毫米包括约217.18个像素,检测的圆度提高0.04%以上,标准差减少不少于7.27%,因此可更准确获取微小轴承内径的亚像素边缘。
For the inner diameter measurement of micro bearing,the machine vision technology is of great prospect for non-contact measurement.Each edge pixel of bearing can be divided into two parts,and the gray value of the edge pixel can be obtained by statistical projection on its both sides with area weighted mean.Taking the rectangular neighborhood of the pixels whose gray values are processed with Gauss weighted method,precision of sub-pixel edge can be improved according to Taylor formula.Experimental results show that can one millimeter consists of approximately 21 7.1 8 pixels,and the roundness increases by 0.04%,and the standard deviation decreases by 7.27%,Thus,more precise subpixel edge for micro bearings can be obtained with the proposed method compared with the linear fitting.