摄像机畸变中心及畸变系数标定方法

Camera Distortion Coefficient and Calibration Method of Distortion Center

针对传统针孔模型下摄像机畸变校正方法应用场景受限,和基于图像特征的畸变校正方法中将图像中心作为畸变中心造成标定精度不高的问题,提出一种基于角点共线特征的畸变中心及畸变系数标定方法。利用过光学中心直线畸变最小的原理,设计迭代算法搜寻到畸变程度最小的两条直线,解算得到精确的畸变中心,进行畸变系数的标定。采用Levenberg-Marquardt算法和多项式畸变模型,利用整幅棋盘格角点拟合的直线特征建立非线性优化模型,标定出一阶径向畸变系数,仿真标定误差在e^(-12)~e^(-11)量级。通过实验验证了畸变中心和畸变系数标定算法的可靠性和准确性,实验结果表明,无论是FA镜头,还是Scheimpflug镜头和广角镜头的畸变,算法均能满足畸变校正的需求。

In order to solve the problem that the traditional camera distortion correction method uses the image center as the distortion center in the distortion correction of image features,which leads to low calibration accuracy,a distortion center calibration method based on the collinear feature of corner points is proposed.By using the principle of the minimum distortion of the straight line in the center of the image,the accurate distortion center is obtained without affecting the calibration of distortion coefficient.Using Levenberg-Marquardt algorithm and polynomial distortion model,a nonlinear optimization model was established based on the linear features of the whole checkerboard corner fitting,and the first-order radial distortion coefficient was calibrated.The calibration error of simulation was in the order of e-09~e-08.The reliability and accuracy of the calibration algorithm for distortion center and distortion coefficient are verified by experiments.The experimental results show that the algorithm can meet the requirements of distortion correction for both FA lens,Scheimpflug lens and wide-angle lens.