Figure 1. The top row: (a) a polyhedral sphere; (b) the sphere with a uniform noise added; (c) Laplacian smoothing develops unnatural deformations; (d) smoothing by the Taubin method converts high-frequency surface oscillations into low-frequency waves; (e) smoothing by the mean curvature ow increases the mesh irregularity; (f) smoothing according to a method proposed in this paper, see Section 3. The bottom row: (a) the Stanford bunny; (b) the bunny with a uniform noise added; (c) Laplacian smoothing with a number of iterations chosen to achieve `a good looking' result: extra iterations will lead to oversmoothing; (d) Taubin smoothing reduces high-frequency surface oscillations but enhances low-frequency oscillations: extra iterations will lead to enhancing of surface wrinkles; (e) smoothing by the mean curvature ow with a number of iterations chosen to achieve `a good looking' result: extra iterations will lead to oversmoothing; (f) smoothing according to a method proposed in this paper: the smoothing process slows down automatically and extra iterations will produce almost the same appearance, see Section 4. - "Polyhedral surface smoothing with simultaneous mesh regularization"
Full article: Tracking Critical Points on Evolving Curves and Surfaces
Contour-parallel tool path generation for three-axis mesh surface machining based on one-step inverse forming - Jinting Xu, Yujin Wang, Xiangkui Zhang, Shuyu Chang, 2013
Energies, Free Full-Text
PDF) A new feature-preserving mesh-smoothing algorithm
Mathematics, Free Full-Text
Non-iterative Global Mesh Smoothing with Feature Preservation
Unstructured surface mesh smoothing method based on deep reinforcement learning
PDF) Polyhedral surface smoothing with simultaneous mesh regularization
Nanomaterials, Free Full-Text
Mesh Denoising Using Quadric Error Metric - Universidade Federal
Full article: Tracking Critical Points on Evolving Curves and Surfaces