单目平行线约束下的空间点坐标恢复  

Extraction of 3-D coordinates using parallel lines in monocular images

作　　者：张元元[1] 吴晓娟[1] 李秀媛[1] 阮秋琦[2] 

机构地区：[1]山东大学信息科学与工程学院,山东济南250100 [2]北京交通大学信息科学研究所,北京100044

出　　处：《智能系统学报》2010年第1期24-29,共6页CAAI Transactions on Intelligent Systems

基　　金：国家自然科学基金资助项目(60675024)

摘　　要：提出一种从单目像机拍摄的二维投影图像中恢复空间点三维坐标的简化算法.假定摄像机与空间中的一对平行线的位置关系已知,同时摄像机坐标系下平行线所在平面中的任意空间点在像平面中的投影也已知,通过射影几何可得图像坐标到目标点三维空间坐标的转换系数,即坐标转换因数,从而估计出三维空间点的坐标.实验中通过对人体行走轨迹进行估计来测试算法性能,结果表明算法简单有效,计算开销小.A method was proposed to simplify the extraction of 3-D coordinates from spatial points in 2-D images captured by a monocular camera.Assume that two parallel lines could be known in the image,and also that their relative positions with respect to the video camera were known.In the camera coordinate system,the projection of any spatial point in the plane determined by parallel lines to the image plane of the camera could be measured.By applying the knowledge of projection geometry,a coordinate conversion factor （CCF） for the given point was then calculated;these were used to estimate 3-D spatial coordinates.Estimations of the paths of people walking were then done to analyze the effectiveness of this method.Results showed that the proposed algorithm is simple and effective,and has low computational costs.

关 键 词：三维重建 单目摄像机 平行线约束 坐标转换因数 坐标恢复 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]