多视角下数据的配准

Registration of Measuring Data under Different Viewpoints

提出了采用最靠近点的迭代 (IterativeClosestPoint,ICP)算法 ,用于完成多视角下数据的配准 ,并采用四元数法来求解曲面配准所需的旋转及平移阵 .在应用上述方法中 ,根据测量数据的特点采用了均匀参数化法 ,并以此为基础利用线性插值法建立了曲面的初始参数模型 .在最靠近点的迭代算法中 ,提出了一种新的判断不同视角下对应点集间有效对应点的方法 ,该方法简便、快速 ,具有实用价值 .

The iterative cloest point (ICP) algorithm is used in this paper to the registration of measuring data under different view points and the rotation and translation matrix are acquired by quatemion method. The parametenization of the measured points is even in terms of the nature of them. The initial paprameter mathematical model is set up with linear interpolation based on the even parameterization. A new way to determine the corresponding point set for registration is put forward, which is more effcient than several well-known ways. Being used in reality, this method runs simply and quickly.