University of Bergen | Faculty of Mathematics and Natural Sciences | Department of Informatics | Visualization Group
Visualization
You are here: Department of Informatics > Visualization Group > Publications > Parulek13Fast
 Visualization
 > about
 > team & contact info
 > research
 > publications
 > projects
 > teaching
 > seminars
 > resources
 > network
 > events
 > links

Fast Blending Scheme for Molecular Surface Representation

Julius Parulek, Andrea Brambilla

ARTICLE, Visualization and Computer Graphics, IEEE Transactions on, Dec., 2013

Abstract

Representation of molecular surfaces is a well established way to study the interaction of molecules. The state-of-theart molecular representation is the SES model, which provides a detailed surface visualization. Nevertheless, it is computationally expensive, so the less accurate Gaussian model is traditionally preferred. We introduce a novel surface representation that resembles the SES and approaches the rendering performance of the Gaussian model. Our technique is based on the iterative blending of implicit functions and avoids any pre-computation. Additionally, we propose a GPU-based ray-casting algorithm that efficiently visualize our molecular representation. A qualitative and quantitative comparison of our model with respect to the Gaussian and SES models is presented. As showcased in the paper, our technique is a valid and appealing alternative to the Gaussian representation. This is especially relevant in all the applications where the cost of the SES is prohibitive.

Published

Visualization and Computer Graphics, IEEE Transactions on

  • Pages: 2653–2662
  • Event: Vis2013
  • Date: Dec. 2013

Media

  • paper
  • additional material
  • Click to view
  • Click to view

BibTeX

@ARTICLE{Parulek13Fast, 
 author={Julius Parulek and Andrea Brambilla}, 
 journal={Visualization and Computer Graphics, IEEE Transactions on}, 
 title={Fast Blending Scheme for Molecular Surface Representation}, 
 month = {Dec.},
 year={2013}, 
 event={Vis2013}, 
 pages={2653--2662}, 
 abstract={Representation of molecular surfaces is a well established way to study 
 the interaction of molecules. The state-of-theart molecular representation is the
 SES model, which provides a detailed surface visualization. Nevertheless, it is 
 computationally expensive, so the less accurate Gaussian model is traditionally 
 preferred. We introduce a novel surface representation that resembles the SES and 
 approaches the rendering performance of the Gaussian model. Our technique is based 
 on the iterative blending of implicit functions and avoids any pre-computation. 
 Additionally, we propose a GPU-based ray-casting algorithm that efficiently visualize 
 our molecular representation. A qualitative and quantitative comparison of our 
 model with respect to the Gaussian and SES models is presented. As showcased in 
 the paper, our technique is a valid and appealing alternative to the Gaussian 
 representation. This is especially relevant in all the applications where the cost 
 of the SES is prohibitive.}, 

}






 Last Modified: Jean-Paul Balabanian, 2014-06-19