Octree Subdivision Using Coplanar Criterion for Hierarchical Point Simplification

This study presents a novel rapid and effective point simplification algorithm based on point clouds without using either normal or connectivity information. Sampled points are clustered based on shape variations by octree data structure, an inner point distribution of a cluster, to judge whether these points correlate with the coplanar characteristics. Accordingly, the relevant point from each coplanar cluster is chosen. The relevant points are reconstructed to a triangular mesh and the error rate remains within a certain tolerance level, and significantly reducing number of calculations needed for reconstruction. The hierarchical triangular mesh based on the octree data structure is presented. This study presents hierarchical simplification and hierarchical rendering for the reconstructed model to suit user demand, and produce a uniform or feature-sensitive simplified model that facilitates rapid further mesh-based applications, especially the level of detail.