机器视觉精密测量系统研究

微小孔系的精密测量用传统的手段是难以实现的。以机器视觉为基础,结合计算机图像处理技术可完成这类测量任务。介绍了基于机器视觉的孔系精密测量系统的硬件组成和图像处理算法。以所提出的识别弦端点来确定图像圆心坐标的算法对500根弦的中点坐标进行滤波处理,可使识别误差达到亚像素级。圆心坐标在X和Y方向的标准差均小于4μm,孔系间角度测量标准差小于3″。

Study on precision measure system based on machine vision

It is difficult to achieve the precision measurement of the position of the micro holes with traditional methods. Based on machine vision and with the help of image processing, this work can be done finely. The hardware structure and the image manipulation algorithm of the holes position precision measure system are described based on machine vision. By the provided algorithm, the center coordinates of circle in the image is recognized by identifying the endpoint of chord, and by the filter processing on the center of 500 chords, it can make the identifying error lower than the distance of image pixel, the standard deviation is less than 4μm in both X and Y direction of the center coordinates of circle, and the angel standard deviation is less than 3″ among the holes.