一种标定相机内方位元素的迭代算法

为优化自标定技术解算内方位元素数学模型，阐述了一种利用自标定技术进行相机内方位元素标定的迭代算法。内方位元素结果优化采用牛顿—拉夫逊迭代算法，在保证计算效率的同时能够获得较高的精度。B双空间几何理论建立于正方形在物、像两空间的射影几何关系，利用二维棋盘格标定图像中点、线、面间固有的几何关系得到迭代算法初值；采用包含3个径向和2个切向畸变系数的畸变模型能够更好地表达镜头畸变情况；标定方法操作过程简单、实验误差小。为与张氏标定法进行比较，对使用张氏标定法的图像数据进行实验，得到与其相同标定精度的实验结果，验证了基于B双空间几何法的自标定方法的正确性和可行性。

A Iterative Method to Solve the Intrinsic Parameters of Camera Calibration

To optimize the math model of using self-calibration technology to solve camera intrinsic parameters, an iterative method is presented to solve the camera intrinsic parameters, based on self-calibration technology. The optimization of intrinsic parameters uses Newton-Raphson iterative method, which ensures the computing efficiency and high accuracy. B-dual-space geometry theory is established in the projective geometry relationship of squares in object and image spaces. The initial guess of iterative method is computed on B-dual-space geometry theory which uses the geometry relationship about points, lines and planes in chessboard images. The distortion model is used which includes 3 radial distortion coefficients and 2 tangential distortion coefficients, to indicate the lens distortion condition better. This calibration method has the merits of simple operation and high precision. To compare with Zhang Zhengyou’s method, the contrastive experiment uses the same images of Zhang Zhengyou’s method and the result shows the same calibration accuracy, which approves the rightness and feasibility of the self-calibration technology based on B-dual-space geometry theory.