Refined Adaptive Meshes from Scattered Point Clouds

This paper presents a new method for adaptive mesh generation from scattered point clouds. We develop new techniques for extracting geometric features from scattered point clouds. In addition, we present how to incorporate these geometric features into implicit mathematical models for fitting scattered, noisy point clouds. This method is also well suited for large and potentially noisy scattered point clouds. Our numerical experiments show that the new adaptive mesh generation algorithms can improve the accuracy of the mesh generation from computed tomography (CT) images.