基于菱形模板的摄像机标定方法

在传统的基于平面模板的摄像机标定算法和基于3个非线性点在特定运动下的标定算法的基础上,提出了一种基于菱形模板的摄像机标定算法,本算法只需要知道菱形的一个夹角就能标定出摄像机的内参数矩阵。弥补了传统算法在标定时需要知道特征点世界坐标的不足,减少了制定模板世界坐标带来的误差;同时本算法不需要菱形特征点做任何运动,避免了旋转特征点带来的人为误差。该算法的标定精度仅取决于特征点检测的精度,实验结果验证了该文算法的有效性。

Camera calibration method based on rhombus template

This paper presents a method of camera calibration using rhombus template, which is based on the research of traditional template-based camera calibration and calibration method with three non-collinear points under special motions. The intrinsic parameters of the camera can be calculated with an angle of the rhombus, avoiding the disadvantage of the traditional method which needs to know the world coordinates of feature points. It reduces the error caused by the world coordinates, and the feature points do not need to do any motions, avoid the human error caused by rotating the feature points. The accuracy of the proposed algorithm depends only on the accuracy of feature points detection, the experimental results show that the proposed method is effective.