Peter Praxmarer
GUP, Joh. Kepler University Linz,
Altenbergerstr. 69, A-4040 Linz, Austria/Europe,
praxmarer@gup.uni-linz.ac.at,
http://www.gup.uni-linz.ac.at/

"Semiautomatic Generation of Transfer Functions through Grid-based
 Parameter Studies"

Direct volume rendering represents a standard method for visualizing
three-dimensional data of considerable size. The transfer function
maps the data to be visualized onto colors and opacities. The
resulting image supports the user's comprehension of the displayed
data.

In this talk an approach for obtaining suitable transfer functions
for presenting astrophysics data is described. A parameter study on
the grid is used for rendering representations of different transfer
functions. The transfer functions are iteratively improved by using
a genetic algorithm, while the user is kept in  the loop to judge
the quality of the rendering results.