Sophisticating Volume Visualization Based on Analysis of 3D Field Topology

Abstract

Volume visualization is a key to understanding complicated inner structures of large-scale volume datasets. Indirect volume rendering (IVR) and direct volume rendering (DVR) are the representatives of such volume visualization techniques; to clarify the inner structures, the IVR extracts isosurfaces from the volume dataset whereas the DVR renders the overall dataset translucently. Although these two techniques have some potentials to generate comprehensible images, they still suffer from selections of related visualization parameters that allow us to emphasize features involved in the target volumes. This paper therefore presents a method of semi-automatically finding appropriate visualization parameter values by analyzing the topological structures of volume datasets. Also presented is another method which enables us to visualize volume datasets having complicated space cofigurations of isosurfces such as mutual inclusion relationships. Finally, a current version of our sophisticated volume viewer is reported. Several application examples are used to demonstrate the feasibility of the present methods.