摘要
根据棋盘格图像的特点,提出了一种新的棋盘格角点检测方法.对图像进行边缘提取后,得到两组相互垂直的直线组,通过Hough变换求出直线组的代数方程,采用代数方法求两组直线之间的交点,确定角点的初始位置.对每一初始角点邻域采用灰度变化检测方法进行精确的角点定位.该方法既保证了角点检出的正确性,又提高了角点坐标的检出精度.试验结果表明,该方法具有较好的效果和较强的鲁棒性.
According to the features of chess board, a new method for detecting the corner of the chessboard was proposed. After edges detected, two groups of orthogonal lines were obtained. In order to get the equations of the line groups, the Hough translation was used to process the image. By solving the equations that obtained from the translation results, through points of intersection, the coarse corner positions were obtained. Then intensity vary detector was used to get the accurate corner coordinates. The experiment results show that the algorithm is efficient and robust.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2005年第9期796-799,共4页
Transactions of Beijing Institute of Technology
基金
教育部博士学科点专项科研基金资助课题(20030007034)
关键词
角点检测
HOUGH变换
边缘提取
灰度梯度
corner detection
Hough translation
edge detection
intensity gradient
