• DocumentCode
    3062938
  • Title

    Texture Atlas Generation for Inconsistent Meshes and Point Sets

  • Author

    Degener, Patrick ; Klein, Reinhard

  • Author_Institution
    Univ. of Bonn, Bonn
  • fYear
    2007
  • fDate
    13-15 June 2007
  • Firstpage
    156
  • Lastpage
    168
  • Abstract
    In order to compute texture atlases with low stretch and hardly visible texture seams existing texture mapping tools pose high demands on the quality of surface representations like consistent orientation, watertightness, or manifoldness which many models commonly used in day-to-day modeling practice fail to meet. In this paper we propose an approach that bridges the gap between requirements of high quality texture mapping tools and poor mesh connectivity of models used in practice. In the spirit of classical two-part mapping an intermediate proxy surface is created that can be processed by high quality texture mapping tools. The texture signal is transferred to the original geometry using a novel mapping technique. Avoiding a modification of the original geometry typical problems of mesh repairing approaches like approximation errors or feature corruption are circumvented. As our method poses almost no demands on connectivity it can also be applied to point clouds. Contrary to classical two- part mapping, the method requires little user-interaction. Its robustness and quality are demonstrated in several examples.
  • Keywords
    approximation theory; computational geometry; computer graphics; mesh generation; approximation errors; feature corruption; geometry; inconsistent meshes; point sets; texture atlas generation; texture mapping tools; texture seams; Approximation error; Bridges; Computer graphics; Computer science; Design automation; Geometry; Robustness; Signal mapping; Surface texture; Visualization;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Shape Modeling and Applications, 2007. SMI '07. IEEE International Conference on
  • Conference_Location
    Lyon
  • Print_ISBN
    0-7695-2815-5
  • Type

    conf

  • DOI
    10.1109/SMI.2007.35
  • Filename
    4273378