一种应用于视觉测量标定的亚像素级特征点提取算法

在应用视觉测量系统时 ,视觉传感器的标定是必要的步骤 ;当采用标定图案标定时 ,角点图像坐标的提取精度会影响最终标定的结果 ,方法的易操作性也会影响该项技术的应用 ;本文提出一种基于亚像素级的特征点提出方法 ,该方法的主要思想是先采用了自动提取的初始点为引导进行边缘提取和哈夫变换 ,再通过直线拟合求交的方法获得角点的精确坐标

An Algorithm of Subpixel Sorner Detection for Calibration Pattern

Calibration is a necessary step before applying photics sensor to measurement.While applying marker in calibration,accuracy of corner image coordinate will finally affect the calibration result,and operation for calibration will also affect the application of the technique.In the paper,a new subpixel corner detecting method is worked out,the essence of which is to first,apply the corner of automatic detection as the guide in finding the edge and in Hough transformation,and then obtain the accurate corner coordinate through line fitting and finding intersection.