VISUALIZATION OF 3-DIMENSIONAL DATA BY VOLUME RENDERING

We report an implementation of a ray-tracing method for visualizing volumetric data. For evaluating the performance of the implementation, we consider the visualization of two classes of volumetric data. The first consists of the probability density functions of a subset of states of an electron in hydrogen atom and represents the case of visualizing data derived from analytical expressions. The second class of volumetric data is a set of three-dimensional fractals which are generated over regular three-dimensional cartesian grids of various sizes. The motivation for generating and visualizing the fractal data sets is two-fold. Firstly a fractal data set is a good test input to verify the correctness of the implementation. Secondly, we want to model stellar clouds and see how they can be visualized. Clouds are known to be akin to fractal data sets and we find that the volume rendered images of these data sets bear a close resemblance to stellar clouds. Ray-tracing is computationally expensive and we therefore report the CPU timings for a representative set of images. General comments on the utility of this method for visualizing in various disciplines conclude the paper.