A topological comparison of surface extraction algorithms

In many application areas, it is useful to convert the discrete information stored in the nodes of a regular grid into a continuous boundary model. Isosurface extraction algorithms differ on how the discrete information in the grid is generated, on what information does the grid store and on the pro...

ver descrição completa

Detalhes bibliográficos
Autores: Andújar Gran, Carlos Antonio|||0000-0002-8480-4713, Brunet Crosa, Pere|||0000-0001-8406-1975, Fairén González, Marta|||0000-0001-7293-584X, Navazo Álvaro, Isabel|||0000-0001-6298-1463, Vinacua Pla, Álvaro|||0000-0001-8984-4311
Tipo de documento: relatório científico
Data de publicação:2021
País:España
Recursos:Universitat Politècnica de Catalunya (UPC)
Repositório:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglês
OAI Identifier:oai:upcommons.upc.edu:2117/359540
Acesso em linha:https://hdl.handle.net/2117/359540
Access Level:Acceso aberto
Palavra-chave:Three-dimensional imaging
Algorithms
Surface extraction
Marching Cubes algorithm
Topologically consistent isosurfaces
Rectangular grids
Discrete volume models
Topology optimization
Imatgeria tridimensional
Algorismes
Àrees temàtiques de la UPC::Informàtica::Infografia
Descrição
Resumo:In many application areas, it is useful to convert the discrete information stored in the nodes of a regular grid into a continuous boundary model. Isosurface extraction algorithms differ on how the discrete information in the grid is generated, on what information does the grid store and on the properties of the output surface. Recent algorithms offer different solutions for the disambiguation problem and for controlling the final topology. Based on a number of properties of the grid’s grey cells and of the reconstruction algorithms, a characterization of several surface extraction strategies is proposed. The classification presented shows the inherent limitations of the different algorithms concerning global topology control and reconstruction of local features like thin portions of the volume and almost non-manifold regions. These limitations can be observed and are illustrated with some practical examples. We review in light of this classification some of the relevant papers in the literature, and see that they cluster in some areas of the proposed hierarchy, making a case for where it might be more interesting to focus in future research.