基于单张平行六面体照片的摄像机标定方法

Scheme for camera calibration based on single parallelepiped photo

摄像机标定是计算机视觉的重要组成部分。根据射影几何理论建立了基于单张平行六面体照片的摄像机全参数标定方法。通过调和射影及Laguerre定理的推论求解出平面圆环点的共轭虚像,利用平行六面体照片存在的三对圆环点的虚像,构造对无穷远平面绝对二次曲线的像ω的约束,使用辗转相除法求解非线性问题。最后,给出该算法对模拟图像和真实照片的实验。实验结果表明计算结果与真实情况吻合。

Camera calibration is one of the most important parts in computer vision.A scheme for camera calibration based on single parallelepiped photo using nonlinear techniques is presented.The numerical algorithm based on theories of projecting and Laguerre,such as homology,circular points and ω of absolute conic is established,an iterative elimination method is applied to boost the nonlinear problem solving.Simulated and real photo experiments are performed,and calibration results are matched with the true situations.