Visualization enhancement by embedding local and global modeling operations in the rendering process

The recent great increase in the availability of powerful computer systems with large memories and storage devices led the way to using memory expensive modalities. Volume graphics is one of these modalities. It requires large amounts of memory for maintaining, and powerful computers to manipulate. Since volume data sets span a three dimensional space, a resolution increase of a volume by some factor n forces storage increase by a factor of ;This dissertation addresses one aspect of volume rendering. It presents methods for increasing the quality and detail of output images without increasing the complexity of the volume data sets. It achieves this task in two families of methods. The two families provide user dependent and user independent tools for deforming the shape of volume models. The user dependent tool requires involvement of the user for generating local deformations in a volume model. Every single deformation requires a direct specification from the user. The user independent method provides a tool for applying global modeling operations on the entire volume data set at once. In this method, the user has to specify one set of inputs, regardless of the complexity of the model.;Both the user dependent and the user independent techniques embed modeling operations in the rendering process. Other modeling methods modify the volume model directly into a new volume model. The new model is often more complex than the original one. None of the methods described herein require any modifications to the original model. They operate in the rendering phase thus providing the illusion that we modified the model.