面向机加工表面粗糙度的光切显微视觉测量系统

摘要

In order to solve the problem of light cutting microscope′s low measurement accuracy and poor efficiency of the machining surface,the vision measurement system over the surface roughness of light cutting microscope is developed based on the theory of light cutting measurement.A visual sensor is used to get the micro profile images of the machining surface,then the contour sub pixel edge detec-tion algorithm is proposed based on the Zernike moments,and a model which combines the maximum between class variance method and the traditional Zernike moments is established,and the positioning accuracy of the contour sub pixel edge is improved.The adoption of the least square fitting method de-termines the contour datum line.According to the national standard of surface roughness,the mathe-matical model of the arithmetical mean deviation of the profile Ra is established,achieving the ma-chined surface roughness measurement.Test results show that when the surface machining accuracy a-chieves level 7 to 10,the relative error of the measured surface roughness is no more than 5%,the al-gorithm takes about 15ms.The measurement system is accurate,and it improves the efficiency of the measurement.%为解决光切显微镜测量机加工表面粗糙度的测量精度不高、效率低的问题，在光切法测量原理的基础上，开发了基于光切显微镜的表面粗糙度视觉测量系统.通过视觉传感器获取机加工表面的微观轮廓图像，提出了基于 Zernike 矩的轮廓亚像素边缘检测算法，建立了最大类间方差法与传统 Zernike 矩相结合的模型，提高了轮廓亚像素边缘点的定位精度.采用最小二乘拟合法确定了轮廓基准中线，根据表面粗糙度的国家标准建立了轮廓算术平均偏差 Ra 的数学模型，实现了机加工表面粗糙度的测量.结果表明，机加工表面精度等级在7~10级时，所测量表面粗糙度的相对误差不超过5%，算法耗时约15ms.该测量系统具有较好的精度和实时性，提高了测量效率.