Please use this identifier to cite or link to this item:

Quality enhancement and relation-aware exploration pipeline for volume visualization

Authors Chan, Ming-Yuen
Issue Date 2009
Summary With the advance of data acquisition methods, various kinds of high quality volumetric datasets are now available for medical and scientific purposes. Due to the explosive increase in size and complexity of the data, the analysis tasks involved pose difficult problems. Volume visualization provides an effective solution which aims at delivering insights of the underlying details or concepts by the means of computer graphics for creating useful views on the data. In fact, effective visualization requires the support of intuitive and interactive exploration techniques as well as proper rendering and presentation of the data in order to facilitate the users in acquiring insights and useful information from the volumes. This thesis focuses on the improvement of the rendered image quality and the exploration pipeline to facilitate the visualization process. In this thesis, we first present two exploration techniques for visualizing volumetric data. Instead of performing tedious manipulation on the volumes, we suggest a quality-driven camera path planning method and a relation-aware visualization pipeline to allow semi-automatic exploration of volumes. Volume features and high-level spatial relations are considered to determine the proper views on the volumes and our objective is to reveal this important information using the proposed visualization techniques. In the second part, the issues of rendered image quality are discussed and we propose two image quality enhancement approaches to ensure that the features in the volume are faithfully presented in the rendered images and correct perception of the volumes is obtained by viewers. An adaptive enhancement framework is also proposed to adjust the rendering parameters for obtaining optimal rendered images. To demonstrate the effectiveness of our solutions, experiments are conducted on different kinds of datasets and the results are evaluated and discussed.
Note Thesis (Ph.D.)--Hong Kong University of Science and Technology, 2009
Language English
Format Thesis
Access View full-text via DOI
Files in this item:
File Description Size Format
th_redirect.html 343 B HTML
Copyrighted to the author. Reproduction is prohibited without the author’s prior written consent.