基于Kruppa方程的相机分步自标定方法  

Step-by-step self-calibration algorithm of digital camera based on Kruppa equations

作　　者：顾晓东[1] 王晓明[1] 刘健[1] 

机构地区：[1]大连理工大学机械工程学院现代制造研究所,辽宁大连116024

出　　处：《大连理工大学学报》2003年第1期82-85,共4页Journal of Dalian University of Technology

基　　金：国家自然科学基金资助项目(59805001).

摘　　要：提出一种运动变焦相机分步自标定的方法,即事先标定相机的纵横比和主点,在假设像素为矩形及主点为图像中心的情况下,基于Kruppa方程线性求解焦距.算法中只需输入基础矩阵,不必预先进行任何类型的投影阵分解或投影束校正.线性求解焦距避免了非线性法的不稳定性.实际数值实验表明该方法简单实用.A novel technique for self\|calibration of cameras with varying focal length is introduced. The aspect ratio and principal point can be estimated beforehand because they are unchangable during camera′s movements, provided that zero\|skew and the principal point are at the center of the image, an linear solution is derived from using the Kruppa equations. The input of the algorithm is a set of fundamental matrices, and it is not necessary to perform any projective factorization or projective bundle adjustment. An step\|by\|step linear algorithm which can avoid unstability of nonlinear problem has been developed. Experiments with real data synthetized have shown that the technique is simple and useful.

关 键 词：分步自标定方法 基础矩阵 KRUPPA方程 摄影测量学 焦距 变焦相机 计算机视觉 

分 类 号：TB852.1[一般工业技术—摄影技术]