Noise Removal from 3D Meshes: An Efficient Approach

Abstract

The current methods for mesh denoising focus on models where all the vertices in the mesh are corrupted with low noise amplitudes.Some of these prevailing techniques would not provide the facility of denoising even when a small percentage of the vertices are corrupted with noise of very high amplitude. To solve the problem,we propose an efficient mesh denoising algorithm for such models in which vertices are corrupted with both high as well as low noise amplitudes. This proposed algorithm is divided into two phases.In the first phase, the corrupted vertices are detected by means of a filter called Localised Coordinate Variation filter. In the second phase, the corrupted vertices are abolished. The results revealed in this paper depicts the claims as mentioned above.