基于机器视觉的立铣刀磨损检测方法研究

Research on wear detection method of endmill based on machine vision

针对传统立铣刀视觉检测流程中图像定位及利用圆弧拟合边缘不精确的问题,提出一种垂直分布的双镜头视觉检测方法,开发了一套加工中心立铣刀机器视觉检测系统。在自适应阈值分割及边缘检测算法的基础上,通过筛选及合并共线轮廓,拟合直线获取端面刀刃偏转角度,驱动电动机精确定位后采集铣刀侧面图像。基于立铣刀投影几何模型复原侧面轮廓,在侧面摄像头所获取的图像上拟合原轮廓曲线,结合区域求差算法计算出磨损参数及位置信息。实验表明,此检测方法可以精确拟合圆周刃轮廓,提取磨损缺陷区域,检测精度达到0.01 mm,实现了加工中心立铣刀在位检测。

Aiming at the problem of image positioning in the visual inspection process of traditional end mill and the inaccuracy of the arc fitting edge,a vertical distributed dual-lens vision detection method is proposed,and a machine vision inspection system for machining center end mill is developed.On the basis of adaptive threshold segmentation and edge detection algorithm,by sifting and merging the collinear contours,the straight line is fitted to obtain the deflection angle of the end face,and the motor is accurately positioned to collect the flank image of the milling cutter.The flank contour is restored based on the end mill projection geometry model,the original contour curve is fitted on the image acquired by the flank camera,and the wear parameter and position information are calculated by the regional difference algorithm.Experiments show that the detection method can accurately fit the contour of the circumferential edge and extract the wear defect area.The detection accuracy reaches 0.01 mm,and the in-situ detection of the machining center end mill is realized.