摘要
论述由位于正交平面上的特征点对集合复原无穷远单应性矩阵和摄像机内参数的新算法.通过对无穷远单应性矩阵的特征分解,计算5参数摄像机内参数阵.提出3种求解极点的方法,并对它们的效果进行对比.实验表明,该算法可行且具有较好鲁棒性.
A new algorithm is presented to determine infinite homography matrix and camera intrinsic matrix parameters under one camera motion set from the matched-point sets on two orthogonal planes. The camera intrinsic matrix with five parameters is obtained through eigen value decomposition of the infinite homography matrix. Based on the different methods obtaining epipole, three techniques are provided and compared. Experiments demonstrate that the new algorithm is feasible and robust.
出处
《上海海事大学学报》
北大核心
2006年第4期80-83,共4页
Journal of Shanghai Maritime University
基金
上海市高等学校科学技术发展基金(01G02)
关键词
单应性矩阵
极点
正交平面
摄像机自标定
homography matrix
epipole
orthogonal plane
camera self-calibration