A feature-preserving mesh denoising filter for 3-D printers

© 2016 IEEE. A feature-preserving mesh denoising filter is proposed in the paper for 3-D printers. The whole algorithm contains two steps: non-iterative normal filtering and iterative vertex updating. The basic idea of normal filtering is to minimize the angles between face normals in meshes in flat region. The proposed filter was tested and compared with other mesh denoising filters on simulated noisy mesh models as well as real mesh models. The performance of filters were evaluated based on the edge features from the filtered meshes. The proposed filter is very simple to accomplish, and it shows good performance in both noise removal and edge preservation. We also printed the original noisy mesh and the filtered versions using a 3-D printer and visually checked the quality of the filtered meshes obtained by different filters and we found that the meshes filtered by the proposed filter has the best visual quality