将坐标测量机与单目视觉测量系统组合起来, 提出了一种触发式测头与视觉测头联合标定的方法。选用内孔径为4 mm的光面环规作为标定件, 并利用转台、精密滑台配合视觉系统进行环规成像面的调整; 然后, 利用视觉测量系统提取环规内孔圆心图像坐标, 包括设置图像的感兴趣区(ROI)、二值化、形态学去噪、亚像素边缘检测和利用最小二乘法提取圆心的图像坐标。通过拟合内孔圆柱面找到轴线, 并利用接触传感系统提取相同位置圆心仪器坐标下坐标; 最后, 采用最小二乘法求解相关方程, 完成联合标定。试验验证显示, 该标定方法的标定精度高于6 μm, 优于已有的标定方法, 能够满足测量系统联合测量的精度需要。

In combination of a coordinate measuring machine and a monocular vision measuring system, a multi-sensor joint calibration method was proposed by a flip-over probe and a vision probe. A 4 mm smooth ring gauge was used as a calibration part and a rotary table, a precise sliding table and a vision system were taken to adjust the imaging plane of the gauge. Then, the vision measurement system was used to extract the inner hole center image coordinate of the ring gauge, involving in setting the Region of Interest(ROI), binarization, morphologic denoising, sub-pixel edge detection and the extraction of cycle center image coordinates by the least square method. Furthermore, by fitting the inner cylinder of the ring gauge to obtain an axis line, the coordinate under the instrument coordinates of the same point was extracted by a trigger probe. Finally, joint calibration was completed by solving the equation using least square method. The experiment results show that calibration error is within 6 μm, which is better than that of existed other calibration methods and can meet the need of joint measurement for measuring accuracy.