A Surface Reconstruction Algorithm based on Octree and Optimized Local Smooth Degree

摘要

Surface reconstruction algorithm is a research issue in the field of computer graphic, and it has wide application prospects in the fields of 3D geographical information system, computer-aided design and manufacture, reverse engineer. The paper proposes a region-growing algorithm for surface reconstruction. Firstly, split the scattered point cloud to child node of specified threshold size using octree; secondly, reconstruct local surface based on Delaunay regulation in child node; then join the local surface and optimize them through local smooth degree optimization standard and adjust the entire surface normal; finally the surface reconstruction is completed. The experiment results show that the algorithm can solve the problem of non-uniform sampling points set and cope with boundary, and it is fit for open and closed surface reconstruction. The complexity of the algorithm is stable and the realization of the algorithm is easy.