The Community for Technology Leaders
Green Image
Issue No. 01 - January (2012 vol. 18)
ISSN: 1077-2626
pp: 160-172
Bernd Hamann , University of California, Davis
Scott E. Dillard , Pacific Northwest National Laboratory, Richland
Mario Hlawitschka , Universität Leipzig, Leipzig
Sohail Shafii , University of California, Davis
ABSTRACT
Scientific data sets generated by numerical simulations or experimental measurements often contain a substantial amount of noise. Smoothing the data removes noise but can have potentially drastic effects on the qualitative nature of the data, thereby influencing its characterization and visualization via topological analysis, for example. We propose a method to track topological changes throughout the smoothing process. As a preprocessing step, we oversmooth the data and collect a list of topological events, specifically the creation and destruction of extremal points. During rendering, it is possible to select the number of topological events by interactively manipulating a merging parameter. The result that a specific amount of smoothing has on the topology of the data is illustrated using a topology-derived transfer function that relates region connectivity of the smoothed data to the original regions of the unsmoothed data. This approach enables visual as well as quantitative analysis of the topological effects of smoothing.
INDEX TERMS
Volume visualization, visualization techniques and methodologies, smoothing.
CITATION
Bernd Hamann, Scott E. Dillard, Mario Hlawitschka, Sohail Shafii, "The Topological Effects of Smoothing", IEEE Transactions on Visualization & Computer Graphics, vol. 18, no. , pp. 160-172, January 2012, doi:10.1109/TVCG.2011.74
104 ms
(Ver 3.1 (10032016))