Transmittance-based Extinction and Viewpoint Optimization
Computer Graphics Forum (Proc. EuroVis) 2024
Friedrich-Alexander-Universität Erlangen-Nürnberg
Abstract
A long-standing challenge in volume visualization is the effective communication of relevant spatial structures that might be hidden due to occlusions. Given a scalar field that indicates the importance of every point in the domain, previous work synthesized volume visualizations by weighted averaging of samples along view rays or by optimizing a spatially-varying extinction field through an energy minimization. This paper introduces an approach to jointly optimize the extinction field and the viewpoint, and we demonstrate that viewpoints can be found that achieve better visibility, higher contrast, and require fewer modifications of the original volume data than when using predefined, static viewpoints. Our approach is based on a novel formulation of the transmittance-based visibility optimization, which incorporates view-dependent importance and opacity transfer functions, and uses modern automatic differentiation to compute the gradients of the energy with respect to the camera parameters. In addition, we propose an interactive visualization technique that allows domain experts to explore the optimization results in real-time.
Interactive
An interactive comparison of extinction optimization methods on the EARTH MANTLE, VISIBLE HUMAN, HEPTANE FLAME, and ROTATING MIXER data set. Viola et al. [VKG04] and Marchesin et al. [MMD10] did not include shadow optimizations and are therefore compared without shadowing. In addition, Ament et al. [AZD17] with and without shadowing were compared with our method. Previous approaches ([VKG04], [MMD10]) provide limited depth perception. Ament et al. [AZD17] has global parameters, making it difficult to bring out differently important structures throughout the domain. Some parts might even be lost entirely. In contrast, our method maximizes the visibility successfully everywhere.
BibTeX
@article{Himmler24EuroVis,
author = {Himmler, Paul and G{\"u}nther, Tobias},
title = {Transmittance-based Extinction and Viewpoint Optimization},
journal = {Computer Graphics Forum (Eurographics Conference on Visualization)},
volume = {43},
number = {3},
year = {2024},
publisher = {Eurographics},
address = {Odense, Denmark},
}Acknowledgements
We thank Gábor Janiga from the University of Magdeburg for providing the ROTATING MIXER. The EARTH MANTLE was simulated by Anna Gülcher from the NASA Jet Propulsion Laboratory. The VISIBLE HUMAN data set was provided by Peter Ratiu via The National Library of Medicine, Bethesda, MD. The HEPTANE FLAME data set was released by the University of Utah Center for the Simulation of Accidental Fires and Explosions. This work was supported by DFG grant no. GU 945/3-1.