On Volumetric Shape Reconstruction from Implicit Forms

Abstract

International audienceIn this paper we report on the evaluation of volumetric shape reconstruction methods that consider as input implicit forms in 3D. Many visual applications build implicit representations of shapes that are converted into explicit shape representations using geometric tools such as the Marching Cubes algorithm. This is the case with image based reconstructions that produce point clouds from which implicit functions are computed, with for instance a Poisson reconstruction approach. While the Marching Cubes method is a versatile solution with proven efficiency, alternative solutions exist with different and complementary properties that are of interest for shape modeling. In this paper, we propose a novel strategy that builds on Centroidal Voronoi Tessellations (CVTs). These tessellations provide volumetric and surface representations with strong regularities in addition to provably more accurate approximations of the implicit forms considered. In order to compare the existing strategies, we present an extensive evaluation that analyzes various properties of the main strategies for implicit to explicit volumetric conversions: Marching cubes, Delaunay refinement and CVTs, including accuracy and shape quality of the resulting shape mesh

VoroCrust: Voronoi Meshing Without Clipping

“…Another limitation is the requirement that the input triangulation is a faithful approximation of the domain. This inhibits the application of this approach to implicit forms [Wang et al 2016], noisy inputs [Mederos et al 2005], or unclean geometries [Attene et al 2013]. In particular, the algorithm does not fill holes or undesirable cracks in non-watertight inputs [Kumar et al 2008].…”

Tracking-by-Detection of 3D Human Shapes: From Surfaces to Volumes

“…3)? We investigate this direction with a volumetric representation from centroidal Voronoi tessellations that haven shown some recent success in various applications [51], [52], i.e. s is a CVT cell.…”

Multi-view Dynamic Shape Refinement Using Local Temporal Integration

“…Although MC would also work in our case we consider instead a different strategy that addresses some of the limitations of MC: MC is based on a regular discretization of the space and hence dilutes precision inside the shape, unless a specific strategy such as subdivision is applied at the surface; MC is not guaranteed to provide manifold meshes, again unless specific and costly additional steps are performed. In contrast we built on recent works on Voronoi Tesselation [41] showing that better precision can be obtained with discretizations of shapes instead of space. We devise a simple yet efficient version of Voronoi Tesselation that specifically accomodates multi-view capture scenarios.…”