Results

In this chapter we apply the techniques previously described to a variety of ``real world'' volume datasets. In addition, the examples will demonstrate the effects of various adjustments to the algorithm which have been alluded to in previous chapters without illustration. We will analyze the following datasets:

• Turbine Blade: A CT dataset is blurred slightly as a pre-process to make the blade's boundary better match the ideal boundary, and to improve rendered surface quality.

• Engine Block: The effectiveness of two dimensional opacity functions is shown. Two dimensional opacity functions reveal structural details of the dataset not visible with any one dimensional opacity function.

• CT Head: This example illustrates the effect of changing histogram volume size on the quality of the generated opacity functions. Also, renderings made with our two dimensional opacity functions permit a comparison between our method and Levoy's method for boundary visualization.

• Spiny Dendrite (hippocampus): Given a dataset in which the measured boundaries are not ideal, some experimentation with $g_{thresh}$ and the boundary emphasis function is needed in order to create an informative rendering.