大尺寸机械零件的机器视觉高精度测量方法

为实现大尺寸机械零件的高精度视觉测量，研究了基于序列局部图像尺寸特征的测量方法，提出了基于纹理特征的序列局部图像校准技术以解决测量过程中相面旋转引起的局部图像尺寸方向变动问题；提出了图像边缘补偿测量技术以消除实际边缘不能精确定位而对测量精度的影响；论述了基于序列局部图像尺寸特征测量方法的算法实现过程。实验表明，对大尺寸零件应用序列图像测量法进行测量，其相对测量误差在0.012%以内，基本满足机械零件二维尺寸精密自动化测量要求。

Research on High-precision Machine Vision Measurement Method for Large Scale Parts

This paper investigated a vision measurement method based on sequential partial images for large scale machine parts. First, to align the sequential images to an identical coordinates and overcome the problems arising from the relative rotation between the camera and the machine part, a rectifying approach based on texture features was proposed. Then, an edge-pixel compensation technique was put forward to get rid of the effects of the transitional distribution of part image edges on the measurement precision. Finally, the algorithm of sequential image measurement method was presented. The experimental results show that the relative error is less than 0.012% using the method to gauge large scale parts. The precision meets the needs of the precise measurement of 2-D dimensions.