Title: Intuitive transfer function editing using relative visibility histograms
Authors: Luo, Shengzhou
Maji, Subhrajyoti
Dingliana, John
Citation: Journal of WSCG. 2018, vol. 26, no. 1, p. 48-57.
Issue Date: 2018
Publisher: Václav Skala - UNION Agency
Document type: článek
URI: wscg.zcu.cz/WSCG2018/!_2018_Journal_WSCG-No-1.pdf
ISSN: 1213-6972 (print)
1213-6980 (CD-ROM)
1213-6964 (on-line)
Keywords: objemové vykreslování;funkce přenosu;viditelnost;histogramy viditelnosti
Keywords in different language: volume rendering;transfer function;visibility;visibility histograms
Abstract in different language: In this paper, we present an interactive approach for intuitively editing colors and opacity values in transfer functions for volume visualization. We introduce the concept of a relative visibility histogram, which represents the difference between the global visibility distribution across the full volume and the local visibility distribution within a user-selected region in the viewport. From this measure, we can infer what subset of the 3D volume the user intends to select when they click on a region in the 2D rendered image of the data set, and use this to modify relevant parts of the transfer function. We use this selection mechanism for two alternative purposes. The first is to allow output-driven editing of the transfer function, whereby a user can change the opacity values and colors of features without directly having to manipulate the transfer function itself. The second is to extract visually dominant features in any user-selected region of interest, so that the user may individually edit their appearance and then merge these to create new transfer functions. Our approach is lightweight compared to similar techniques and performs in real-time.
Rights: © Václav Skala - UNION Agency
Appears in Collections:Volume 26, Number 1 (2018)

Files in This Item:
File Description SizeFormat 
Luo.pdfPlný text11,28 MBAdobe PDFView/Open

Please use this identifier to cite or link to this item: http://hdl.handle.net/11025/34588

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.