Dual Voronoi Clustering and Remeshing

This paper presents an efficient algorithm for generating multiresolution representations of higher quality by employing Voronoi-Delaunay triangulation. It clusters Voronoi regions on dual polygonal meshes and therefore automatically satisfies the constraint that no more than three Voronoi tiles to share a corner. In addition,it also selects sites under the guidance of curvature distribution in order to capture the geometric features of 3D models. Finally,a resampling strategy combining Loop subdivision and Laplacian smoothing is introduced to enhance the quality of remeshing results. As Voronoi partition is the bottleneck of the algorithm,the adoption of dual polygonal meshes substantially reduces the time for checking the validity of Voronoi partition,hence the algorithm's efficiency.