基于平行直线的摄像机标定方法  

Camera calibration with parallel lines

作　　者：马长正[1] 董明利[1] 祝连庆[1] 吕乃光[1] 

机构地区：[1]北京信息科技大学光电信息与通信工程学院,北京100192

出　　处：《北京航空航天大学学报》2009年第10期1210-1213,1219,共5页Journal of Beijing University of Aeronautics and Astronautics

基　　金：国家自然科学基金资助项目(50675015;50475176);北京市属市管高等学校人才强教计划资助项目(PXM2007_014224_044674)

摘　　要：针对视觉测量中摄像机镜头畸变对精度的影响问题,提出一种基于平行直线求解镜头畸变系数及摄像机标定的方法.考虑到空间直线在理想摄像机模型下的成像仍然是直线的特点,根据畸变对直线的影响,利用径向畸变的数学模型和参数分解的方法对径向畸变系数进行线性求解;仅存在切向畸变的点和理想点到主点距离相等,采用非线性拟合直线确定理想点的方法求解出切向畸变系数.对标靶点进行畸变校正,计算出摄像机内外方位参数.实验结果表明,该方法得到了准确可靠的标定参数.To solve precision problem of the camera lens distortion in vision measurement, a new method for nonlinear distortion correction and camera calibration was presented. Considering the fact that a camera followed the perspective camera model if and only if the projection of every 3D line in space onto the camera plane was a line, the coefficients of radial distortion were calculated with the radial anamorphic model and linear characteristic. The distances from the points with tangential distortion and the ideal points to the principal point were equivalent, and the coefficients of tangential distortion were obtained with the ideal points gotten by imitating the straight line. The distortion coefficients were used to correct the distortion points, and the param- eters of the camera were estimated. The experimental results show that the accurate and reliable calibration parameters were obtained.

关 键 词：视觉测量 畸变系数 摄像机标定 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]