• DocumentCode
    2681950
  • Title

    A Two Pass Fuzzy Logic-based Approach for Efficient Rendering of Point Sampled Surfaces

  • Author

    Bhakar, Sushil ; Rajagopalan, Ramgopal ; Mudur, Sudhir

  • Author_Institution
    Dept. of Comput. Sci. & Software Eng., Concordia Univ., Montreal, Que.
  • fYear
    2006
  • fDate
    3-6 June 2006
  • Firstpage
    329
  • Lastpage
    334
  • Abstract
    Point sampled surface models are gaining tremendous popularity in 3D gaming and cinema due to the recent advances in 3D object scanning technology. Splatting is the standard technique used for rendering mainly to ensure the absence of holes due to inadequate sampling, and the need for visual smoothness in the final rendered image. Since a pixel may be covered by multiple splats, high fidelity rendering requires correct visibility computations and blending at every pixel. In this paper, we present a two pass fuzzy logic based approach for increasing rendering efficiency. Each splat belongs to pixels in a fuzzy way with probability reducing as the distance from centre of splat is increased. The first pass for visibility computation is a fast one as it is computed at lower resolution. In second pass, we blend the splats which contribute to a pixel. For blending, we use a simple fuzzy rule to find the contribution of each splat. We have carried out a complete implementation of this technique and the results clearly demonstrate the performance gain when rendering large point sampled surface models
  • Keywords
    fuzzy logic; octrees; rendering (computer graphics); 3D object scanning technology; blending; point sampled surfaces; rendering; splatting; two pass fuzzy logic; Computational geometry; Computer graphics; Computer science; Filtering; Fuzzy logic; Image sampling; Motion pictures; Rendering (computer graphics); Software engineering; Spline;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Fuzzy Information Processing Society, 2006. NAFIPS 2006. Annual meeting of the North American
  • Conference_Location
    Montreal, Que.
  • Print_ISBN
    1-4244-0362-6
  • Electronic_ISBN
    1-4244-0363-4
  • Type

    conf

  • DOI
    10.1109/NAFIPS.2006.365431
  • Filename
    4216824