智能三坐标测量机零件位姿识别中的立体匹配

Stereo Vision Matching for Position and Orientation Recognition of Parts on Intelligent CMMs

为提高三坐标测量机零件位姿识别的效率和准确率,提出了一种基于图像质心偏移的立体匹配新方法。将CCD摄像机安装在三坐标测量机的Z轴上,利用三坐标测量机的精确平移特性,构成单摄像机立体视觉系统,在两个不同位置获取被测零件的实时图像,对其进行处理得到图像质心,将图像质心的偏移距离作为约束条件完成立体匹配。该方法避开了极线约束、灰度约束等复杂约束条件,实验表明该方法可以达到较高的匹配精度,实验件的匹配时间为1.818s。

In order to improve the efficiency and accuracy of position and orientation recognization of parts on Three-coordinate measuring machines （CMM）, a new method of stereo vision matching on the basis of displacement of image centroid is proposed. The CCD camera is fixed on the Z axis of CMM, and single camera stereo vision system is composed. Taking full advantage of the precision movement of CMM, real-time images of the part to be mesured are captured on two different positions, and two image centroids are acquired by image processing. The offset distance between two centroids is used as a restriction to accomplish stereo vision matching. The above method avoids using some complicated restriction such as epipolar constraint and gray constraint. The experimental results show this method is of high pitching precision, and the image pitching time of the measured part is 1. 818 s.