Video n. 118

Efficient Visualization of Large Scale Models on Commodity Graphics Platforms

Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno

  • Date: 01/2004
  • Duration: 00:07:50
  • Production: CRS4 - D
  • Master: CD118
  • Encoding: DivX
  • thumbnail

    [ HIGH QUALITY DOWNLOAD]
    165595 KB

    Abstract

    CRS4's research activities are concentrated on the development of enabling technologies in physical modeling, mathematical modeling and computer science in high priority areas recognized as strategic at the regional, national and European levels. The principal applications are in the sectors of energy, environmental science, life sciences and the information society. This video, focusing on the efficient visualization of large scale models on commodity platforms, is one of the many possible examples of research carried out at the center.

    The need for interactively inspecting very large surface meshes, consisting of hundreds of millions of polygons, arises naturally in many application domains, including 3D digital photography, 3D scanning, geometric modeling, and numerical simulation. However, despite the rapid improvement in hardware performance, these meshes largely overload the performance and memory capacity of state-of-the-art graphics and computational platforms. A wide variety of simplification methods and dynamic multiresolution models have been proposed to face the problem, but, unfortunately, none of them is able to perform both scalable simplification and interactive view-dependent visualization of very large meshes without imposing a lossy decimation of the original dataset. This is mainly because current methods, heavily CPU bound, are unable to generate model updates at full GPU speed and to efficiently communicate them to the graphics hardware through preferential data paths. This CPU/GPU gap is doomed to widen, since CPU processing power grows at a much slower rate than GPU's one.

    CRS4, in collaboration with ISTI/CNR, has recently proposed a new breed of solutions for interactive and accurate visualization of very large surface models on consumer graphics platforms. The underlying idea of the proposed methods is to depart from current point- or triangle-based multiresolution models and adopt a cluster-based data structure, from which view-dependent representations can be efficiently extracted by combining precomputed patches. Since each patch is composed of a few thousand elements, the multiresolution extraction cost is amortized over many graphics primitives, and CPU/GPU communication can be optimized to fully exploit the complex memory hierarchy of modern graphics platforms. We have demonstrated the performance of the approach with specialized methods for regional to planetary scale phototextured terrain models (BDAM and P-BDAM techniques), isosurfaces and general surface meshes (Adaptive Tetrapuzzles technique), as well as point-sampled models (Layered Point Clouds technique).

    Related Publications

    thumbnail
    [1] Enrico Gobbetti and Fabio Marton. Layered Point Clouds - a Simple and Efficient Multiresolution Structure for Distributing and Rendering Gigantic Point-Sampled Models. Computers & Graphics, 28(6): 815-826, December 2004. 
    thumbnail
    [2] Enrico Gobbetti and Fabio Marton. Layered Point Clouds. In Marc Alexa, Markus Gross, Hanspeter Pfister, and Szymon Rusinkiewicz, editors, Eurographics Symposium on Point Based Graphics. Pages 113-120, 227. Eurographics Association, June 2004. Conference held in Zurich, Switzerland, June 2-5, 2004. 
    thumbnail
    [3] Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno. Adaptive TetraPuzzles - Efficient Out-of-core Construction and Visualization of Gigantic Polygonal Models. ACM Transactions on Graphics, 23(3): 796-803, August 2004. Proc. SIGGRAPH 2004. 
    thumbnail
    [4] Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno. Interactive Out-of-core Visualization of Very Large Landscapes on Commodity Graphics Platforms. In Olivier Balet, Gérard Subsol, and Patrice Torguet, editors, International Conference on Virtual Storytelling. Volume 2897 of Lecture Notes in Computer Science. Pages 21-29, Springer Verlag, New York, NY, USA, November 2003. Proc. Second International Conference, ICVS 2003, Toulouse, France, November 20-21, 2003, Proceedings. 
    thumbnail
    [5] Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno. Batching Meshes for High Performance Terrain Visualization. In Proceedings Eurographics Italy, September 2003. CDROM Proceedings.
    thumbnail
    [6] Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno. Planet-Sized Batched Dynamic Adaptive Meshes (P-BDAM). In Proceedings IEEE Visualization. Pages 147-155. IEEE Computer Society Press, October 2003. 
    thumbnail
    [7] Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno. BDAM - Batched Dynamic Adaptive Meshes for High Performance Terrain Visualization. Computer Graphics Forum, 22(3): 505-514, September 2003. Proc. Eurographics 2003 - Second Best Paper Award. 

    Related Videos and Multimedia Productions

    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Batched Dynamic Adaptive Meshes for Terrain Rendering - Large Scale Stereoscopic Display Demostration: flight over Mars
    CRS4 Video n. 123 - Date: 11/2004
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Batched Dynamic Adaptive Meshes for Terrain Rendering - City Scale Enhanced Display Demostration: flight over Nice, France
    CRS4 Video n. 122 - Date: 11/2004
    thumbnail
    Enrico Gobbetti and Fabio Marton
    Layered Point Clouds
    CRS4 Video n. 120 - Date: 04/2004
    Presented at the First Eurographics Symposium on Point-Based Graphics, Zurich, Switzerland, June 2-4, 2004.
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Efficient Visualization of Large Scale Models on Commodity Graphics Platforms
    CRS4 Video n. 118 - Date: 01/2004 - Duration: 00:07:50
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Batched Dynamic Adaptive Meshes for Terrain Rendering - Region Scale Demostration: flight over Germany
    CRS4 Video n. 116 - Date: 11/2003
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Batched Dynamic Adaptive Meshes for Terrain Rendering - City Scale Demostration: flight over Nice, France
    CRS4 Video n. 111 - Date: 06/2003
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    Batched Dynamic Adaptive Meshes for Terrain Rendering - Planet Scale Demostration: flight over Mars
    CRS4 Video n. 110 - Date: 06/2003
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    P-BDAM: Planet Sized Batched Dynamic Adaptive Meshes
    CRS4 Video n. 108 - Date: 03/2003
    Presented at the IEEE Visualization 2003 Conference. Conference held in Seattle, WA, USA, October 19-24, 2003.
    thumbnail
    Enrico Gobbetti, Leonardo Spano and Marco Agus
    Hierarchical Higher Order Face Cluster Radiosity for Global Illumination Walkthroughs of Complex Non-Diffuse Environments
    CRS4 Video n. 107 - Date: 02/2003
    Presented at the Eurographics 2003 Conference.
    thumbnail
    Paolo Cignoni, Fabio Ganovelli, Enrico Gobbetti, Fabio Marton, Federico Ponchio, and Roberto Scopigno
    BDAM - Batched Dynamic Adaptive Meshes for High Performance Terrain Visualization
    CRS4 Video n. 106 - Date: 01/2003
    Presented at the Eurographics 2003 Conference.

    See Also

    This research is done in collaboration with the ISTI-CNR Visual Computing Group .