一种新型棋盘格角点检测算法  

A novel corner detection algorithm of chessboard image

作　　者：毕运锋 许钢[1] 江娟娟[1] 吴益红[1] 刘亚梅[1] 

机构地区：[1]安徽工程大学检测技术与节能装置省级重点实验室,安徽芜湖241000

出　　处：《南阳师范学院学报》2016年第3期25-28,共4页Journal of Nanyang Normal University

摘　　要：针对摄像机标定中棋盘格标定板的角点检测,给出了一种基于图像像素点分布的角点检测算法.首先通过判断像素点8邻域范围内相似点个数除去平坦区域点,然后根据棋盘格图像高度对称的特性提取对角线上矩阵,通过此矩阵来判定中心点是否为候选角点,最后通过计算候选角点的算术平均求取最终角点坐标.Aiming at the corner detection for calibration board of chessboard pattern in the camera,a kind of corner detection algorithm has been put forward based on the image pixel distribution. Firstly,the even zone points are removed by estimating the number of similar points around the pixel point 8. Then the upper matrix of diagonal is extracted according to the feature of height symmetry in chessboard pattern image to determine whether the central point is the corner candidate. At the end,through the calculation of the selected corner candidates getting the coordinate in average and the final angle point is obtained. The experimental result shows that the method can solve the testing problem of the chessboard angle points in a better way.

关 键 词：摄像机标定 角点检测 棋盘格 对角线矩阵 

分 类 号：TN911.73[电子电信—通信与信息系统]