A method for dynamic simplification of massive point cloud

Abstract

Aiming at point cloud simplification often loses too many feature details, a simplification method for massive unstructured point cloud is proposed to preserve the geometrical and features. Firstly, the normal vector of the data points is calculated, and the feature points are calculated by position and curvature of the data points. Sharp feature points and edge feature points which are extracted based on gauss map are used to segment the point cloud. Finally, according to the segmentation of the point cloud and the eigenvalue of the constraint points, the point cloud is dynamic simplified. Experimental results show that the proposed method not only preserves the features of the point cloud, but also have the high rate of simplification.