摘要:本文介绍了基于机器视觉的汽车轮毂螺栓测量图像处理技术,数字图像处理技术 主要包括图像的二值化、边缘检测、轮廓和特征点提取及通过特征点构造螺纹的三维 壳体，最后通过标定转化为实际螺纹尺寸。

　　　基于照片恢复空间实体或者场景是机器视觉中的一个重要问题。照片真实客观地 反映了空间实体,可以实现从一组照片上记录的信息来恢复物体的空间拓扑结构。在 由多张视图重建对象物体模型中,一个在参照坐标下的三维点对应的二维投影必定在 连接照相机的光学中心和二维投影点的光线上,从理论上讲,通过一个空间点在不同观 测位置上的二维投影,可以恢复该点的三维坐标。

　　　研究探讨了利用数字图像处理技术进行螺纹几何参数测量的方法,并给出了螺纹 的尺寸测量结果。从理论和实践上证明了该方法的可行性和正确性。

关键词：螺纹，图像处理，机器视觉

Abstract:This article describes the technology based on machine vision measuring wheel bolts image processing, digital image processing includes image binarization, edge detection, contour extraction and planar contour and feature points through three-dimensional shell triangle and so on, and finally through the calibration transformed into the actual thread parameters.

　　　Photo restoration of space-based entity or a machine vision scene is an important issue. Objectively reflect the real picture of spatial entities can be achieved from a group of photos on the recorded information to restore the object space topology. Reconstruction by more than one view of the object model, a three-dimensional reference coordinate point corresponding to the two-dimensional projection must connect the camera's optical center and  two-dimensional  projection  points  of  light  on  the  theory,  through  a  space  point different  observation  position  on  the  two-dimensional  projection,  you  can  restore three-dimensional coordinates of the point.

　　　Study of the use of digital image processing technology thread method to measure geometric parameters, and gives some thread size measurements. The result of theory and practice show that the method is feasible and correct.

Key words: thread, image processing, achine vision