Feature preserving noise removal for binary voxel volumes using 3D surface skeletons

Abstract

Skeletons are well-known descriptors that capture the geometry and topology of 2D and 3D shapes. We leverage these properties by using surface skeletons to remove noise from 3D shapes. For this, we extend an existing method that removes noise, but keeps important (salient) corners for 2D shapes. Our method detects and removes large-scale, complex, and dense multiscale noise patterns that contaminate virtually the entire surface of a given 3D shape, while recovering its main (salient) edges and corners. Our method can treat any (voxelized) 3D shapes and surface-noise types, is computationally scalable, and has one easy-to-set parameter. We demonstrate the added-value of our approach by comparing our results with several known 3D shape denoising methods.