Multi-Volume Visualization and Exploration
Master Degree Thesis
by Stian H. Eikeland
supervised by Helwig Hauser
Abstract
The amount of available volumetric data for scientific data visualization has increased rapidly of the years, calling for new methods to visualize large multi-volume databases. A method for exploring and visualizing multi-volume datasets is presented in this thesis. The work consists of a linked three component setup featuring a metadata overview visualizing, a slice stack visualizing and a focus volume visualization. The system relies on externally provided and internally derived metadata to provide a good overview of multi-volume datasets using a parallel coordinates view. Through brushing and focus+context visualization it is possible to select interesting subsets of the data for further exploration using two linked direct volume rendering views, a slice stack visualizing corresponding slices from selected volumes and a focus volume providing a good detailed view of any selected volume. Through three-way interaction it is possible to efficiently explore and visualize large multi-volume datasets using this setup.