一种改进的三维点集表面重建的区域生长算法

An Improved Region-Growing Algorithm for Surface Reconstruction from 3D Irregular Points

介绍了一种改进的三维点集表面重建的区域生长算法。提出了一种基于三角形任意一边邻域拓扑关系的邻接三角形定位方法,充分利用了法向信息,采用由粗到精的提取策略,减少了候选表面三角形的数量,显著加快了表面重构的整体效率。提出了相应的局部拓扑一致性检测方法,确保了算法的稳健性,解决了传统算法的重叠面片和表面空洞等问题,并且重构的三角网格表面与被采样的物体表面拓扑差别最小。实验结果表明,本文算法可以重构具有任意拓扑的复杂表面。

An improved growing algorithm is presented for surface reconstruction from true three-dimensional points. Firstly, a novel method is designed for neighboring triangle location. This method takes the full advantages of surface normal to extract a preliminary set of candidate triangles. A method for local topological consistency test is proposed to ensure topologically correct reconstruction, the reconstructed surface thus has only small topological difference from the original surface. Experimental results show that the algorithm can efficiently obtain the reconstructed mesh surface with arbitrary tooology.