The Community for Technology Leaders
Green Image
Issue No. 01 - January/February (2006 vol. 26)
ISSN: 0272-1716
pp: 46-55
Kirk Riley , Purdue University
Yuyan Song , Purdue University
Martin Kraus , Purdue University
David S. Ebert , Purdue University
Jason J. Levit , National Oceanic and Atmospheric Administration
Although graphics hardware is well suited to accelerate texture-based volume rendering methods for volume visualization of uniform grids, hardware acceleration is less easily exploited for texture-based volume rendering more complex meshes. Meteorological simulation and measurement data are gridded according to the requirements of the simulation and/or the measurement system. Therefore, an intermediate resampling step is often necessary for texture-based volume rendering, eliminating any potential benefits offered by the original grid structure and often introducing resampling errors. This article presents a novel, texture-based volume rendering architecture applicable to many nonuniform meshes, which avoids the need for resampling or for the visibility sorting typically required for tetrahedral meshes. Advanced lighting methods in texture-based volume rendering improve the perception of 3D structure in the volume. Functional approximations of the mapping from world space to computational space, performed per-fragment on programmable graphics hardware, allows application of these advanced techniques to many important grids in meteorological research at interactive rates.
weather visualization, grid structures, volume rendering, volume visualization

K. Riley, M. Kraus, D. S. Ebert, Y. Song and J. J. Levit, "Visualization of Structured Nonuniform Grids," in IEEE Computer Graphics and Applications, vol. 26, no. , pp. 46-55, 2006.
91 ms
(Ver 3.3 (11022016))