一种改进的单视图相机标定算法  

Single-view Calibrating Algorithm Based on Geometric Constraints in Scene

作　　者：石晶[1] 彭文成[1] 杨明[1] 吴东亚[1] 

机构地区：[1]装甲兵工程学院,北京100072

出　　处：《系统仿真学报》2013年第S1期278-281,共4页Journal of System Simulation

摘　　要：在视景仿真中,仿真场景中通常存在着大量平行性和正交性的几何约束信息。目前,很多算法正是通过利用场景中存在的各种几何约束信息来解算、修正并正确标定摄像机的相关参数,从而能够获得高精度的场景渲染显示效果。即将介绍的算法,就是对其中一种算法进行改进,通过利用场景中存在的三组互相正交的平行线,解算出三组平行线对应方向上的影灭点,继而求解摄像机标定所需的各个参数,并设计了算法实现流程,进行了计算机实现和验证性试验,试验结果比较理想,较好的证明了本算法的有效性和可行性。In scene simulation,there exist plentiful geometric constraints of parallelism and orthogonality.In many other calibrating algorithms,the parameters of the camera were calibrated by the geometric constraints information in scene.In the calibrating algorithm,three groups of orthogonal parallels were chosen,then three vanishing points of parallel directions were calibrated,and the parameters of the camera were calibrated.Algorithm flow which was realized and demonstrated by computer was designed.The result of experiment is expected.The effectiveness and practicability of this algorithm are proved preferably.

关 键 词：相机标定 视景仿真 几何约束 影灭点 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]