中文摘要：

为正确判断点与点之间的邻接关系，结合单站地面激光扫描点云数据的特性，提出利用球面投影对其构建三角网的方法，讨论球面Delaunay构网方法的若干细节。采用球面有向搜索和球面LOP局部优化算法，提高构网时间效率。实验表明本方法能快速、有效地对单站激光扫描点云构网。

英文摘要：

A Triangulated Irregular Network （TIN） generating algorithm is proposed for one station terrestrial laser scanning point cloud. The algorithm is incremental insertion generating algorithm using spherical projection strategy. Previous 3D TIN generating algorithms have problem to get correct neighboring relation between points, so, spherical projection strategy is employed to resolve the problem. The strategy is on the basis of the characteristic of one station terrestrial laser scanning point cloud. Incremental insertion generating algorithm takes less computer space resource, so the algorithm is suit for the data set with large amount of points. Some details of generating spherical Delaunay TIN are described, they include： the data structure of the algorithm; the definition of the projective sphere; generating of initial TIN; the method to judge positional relationship between geometric objects; the method of directed searching for triangle and the spherical local optimization procedure. The processes are optimized to improve the algorithm＇s efficiency. There are two experiments comparing the proposed algorithm with Geomagic Studio to test the performance of the algorithm. The first experiment reveals that the proposed algorithm is more efficient than the Geomagic Studio, and the second experiment shows that the proposed algorithm could get the better results to generate TIN.