Efficient focus+context visual exploration of volume datasets

Abstract

We propose a hierarchical representation of volume data sets based on a user-given definition of features. The model is general and it can be used for various types of feature definitions such as rendering queries, topological skeletons and property-based classification. We use this model to interactively select features of the volume and to render them providing focus+context information. In addition, our model provides a mechanism to index directly the feature voxels, therefore it considerably speeds up rendering.