1、 I 摘 要 视觉是人们感知外部世界的主要途径。随着计算机技术的不断发展，计算机视觉 和立体视觉得到了极大的发展。摄像机标定是机器视觉的基础，因此研究摄像机标定 方法具有重要的意义和实际应用价值。 本文主要研究摄像机标定理论和方法，获取摄像机的有关参数，建立起三维空间 物体与二维图像间的对应关系， 为计算机视觉和立体视觉的下一步研究提供可靠的数 据并打下良好的基础。 文中首先对摄像机标定的基础知识进行了详细的讲解， 然后介绍了几种经典标定 方法。在第四章提出了本文所用的一种较为灵活的方法基于共面点的标定方法。当 精度要求低时，忽略镜头畸变，采用线性方法标定出所有参数。当精度要求高时，可 引入
2、Zhang 畸变模型，利用非线性方法求解畸变系数。在文章的最后，本文利用畸变 系数对图像进行了畸变恢复。首先，分析了图像畸变的原理；其次，通过畸变模型得 到一个二元高阶非线性方程组；最后，利用牛顿迭代法求解方程组，计算得到对应的 无畸变图像上的点，从实现了畸变图像的校正。 在本文中，提出了一种利用 Hough 算法及最小二乘法提取标定点的方法。该算法 比较简单，精度较高。 关键词：计算机视觉 摄像机标定 畸变系数 Hough 算法 II Abstract Human understands the outer world mainly by vision. Now with the devel
3、opment of the computer technology,both computer vision and stereo vision have been developing greatly.Camera calibration is the foundation of the computer vision. Therefore,researh on the camera calibration methods has great important significance of theoretical study and practical value. In this th
4、esis,we research the theory and methods on the camera calibration,and gain the parameters about the camera .With the help of these parameters,we can construct the relationship between the 3D object and the 2D images .Whatsmore,we can provide the reliable data and lay a good foundation for the next r
5、esearch on computer vision and stereo vision. In this thesis,we first introduce the foundational knowledge of camera calibration in detail,and then introduce some classical calibration algorithms.In the four chapter,we have introduced a algorithm which is used in this paper.This is a flexible algori
6、thm using a coplanar target.When we require low accuracy ,we can neglect the lens distortion and linearly solve all the parameters.When high accuracy is required ,we can include Zhangs lens distortion model and solve the distortion coefficient by nonlinear algorithm.In the last of the paper,the deformation of the image have been recovered by putting in distortion coefficient.First of all,the principle of distortion is analysised.Then a dual high-end nonlinear equation is get by the distortio
