A simple camera calibration method based on sub-pixel corner extraction of the chessboard image
A simple camera calibration method based on sub-pixel corner extraction of the chessboard image
复制标题
DOI：
10.1109/icicisys.2010.5658280
复制
发表时间：
2010-12
期刊：
影响因子：
--
通讯作者：
Xingfang Yang;Huang Yumei;Gao Feng
中科院分区：
文献类型：
--
作者：
Xingfang Yang;Huang Yumei;Gao Feng
针对SUSAN算子在检测棋盘角点（正交X角点）等特殊类型角点时的局限性，在SUSAN圆形模板的基础上增加两个环形模板，对SUSAN算子进行了改进。改进的SUSAN算子还包括利用亮度重心对角点进行亚像素定位。此外，根据光学透镜畸变分布的成像特点，提出了一种简单的摄像机标定方法，即先利用图像中心附近的标定点线性标定摄像机的内外参数，然后使用整个视场中的校准点计算畸变系数。实验结果表明，该算法在标定精度和速度上都取得了较好的效果。
In view of the limitations revealed in detecting some special types of corners such as chessboard corners (orthogonal X- corners), the authors of this paper improve the SUSAN operator by adding two ring-shaped masks within the SUSAN circular mask. The improved SUSAN operator also includes sub-pixel positioning of the corners by using the center of the gravity about brightness. Moreover, according to the imaging characteristics of optical lens about distortion distribution, we propose a simple camera calibration method in which the intrinsic and extrinsic parameters of the camera and the distortion coefficients are calibrated separately, that is, the camera intrinsic and extrinsic parameters are linearly calibrated firstly using calibrating points near the center of the image, and then the distortion coefficients are calculated using calibrating points in the whole field of view. The experimental results show that the proposed algorithm obtains good results both in calibrating precision and speed.