摘要
传统的三角网生长法进行点云数据表面模型重建时,搜索第三点耗时太长,导致重建效率很低。采用自适应八叉树划分算法将点云数据分割成相互覆盖的子域,在每个子域内进行三角网格重建,避免网格拼接的过程;采用最大角最小化原则进行三角网格优化;并运用三角面片定向的方法进行网格法向量一致化处理。实验结果表明,该方法极大地提高了表面模型重建的效率,形成的网格质量也很好,能够较好地体现模型的细节特征,鲁棒性好。
In surface model reconstruction with regard to point cloud data, traditional triangulation growth method has very low efficiency because it takes too long time in searching the third vertex. The adaptive octree subdivision algorithm is used in this paper. Point cloud data are divided into subdomains covering each other, and the triangular grids are reconstructed in every subdomain, thus the process of the grid stitching is avoided. The produced triangular grids are optimised using the principle of minimising the maximum angle. A triangular facet orientation method is used for uniformisation of the normal vectors of grids. Experimental results show that the efficiency of the surface model reconstruction is greatly improved by using this method, and the quality of the triangular grids produced are very good as well, the detail features of the model are better reflected, and the algorithm is robust.
出处
《计算机应用与软件》
CSCD
北大核心
2013年第6期83-87,共5页
Computer Applications and Software
基金
教育部留学回国人员科研启动费项目(K314020901)
中央高校基本科研业务费专项资金资助项目(Z109021004)
关键词
点云
表面模型重建
自适应八叉树
三角网生长法
Point cloud
Surface model reconstruction
Adaptive octree
Triangulation growth method