中文摘要：

提出并实现了一种空间点定位方法。该方法仅用一台数码相机围绕目标进行多角度自由拍摄获得多帧图像,采用一套具有身份唯一性的标记点实现多图像间点的稳定匹配;在此基础上根据多视图几何约束,精确计算各次拍摄时的相机姿态和位置;进而求解目标点的三维位置信息,并对结果进行优化。实验表明空间点定位的相对误差优于0.03%。该方法所用设备十分简单,操作方式非常灵活,适用于大型物体多视角测量的拼合、某些情况下的装配检查等,也可直接用于较规则构型物体的三维数据获取与CAD模型重建。

英文摘要：

A flexible 3D target location method is proposed, which is based on a series of digital images taken by one single hand-held CCD camera. A set of coded points is applied to achieve the robust point correspondences between image pairs. The camera position and orientation of every shot are then computed according to the multiple view constraints. Multiple shots with known （estimated） position and orientation actually equal to a multi-nocular setup, from which the 3D coordinates of the un-coded points attached to the object surface can be reconstructed. Finally, a global optimization is used to further refine the result. Experiments show that the relative error for locating 3D points is lower than 0.03%. The proposed method can be used for measuring data registration, interference checking in some assembly scenarios and CAD model reconstruction of objects with relatively regular shape.