Visualization of n-dimensional heterogenous data

Dean's award for outstanding master thesis

This thesis deals with a method of rendering volumetric data, which allows us to combine various visualization techniques while accumulating color and transparency along the ray. The aim of this work is to introduce the reader to the volume rendering on GPU and to show how two-level volume rendering can be used on segmented data. Based on the analysis and suggested design of the solution, an implementation is made. This allows us to compose two newely implemented methods of illustrative rendering with existing methods of direct volume rendering. The resulting vizualization application was tested on six volumetric models with grids of different resolutions, demonstrating what effects can be achieved using different techniques and their various combinations. This provides an insight into the behavior of two-level volume rendering under varied conditions.