|
Even before printing was developed, scientific illustration was
an important means of design depicting facts that are very complex
and hard to put into words. In order to enhance the communication of
essential information illustrations are still used in textbooks
(medicine, biology etc.), construction plans, and similar publications.
Illustration follows the proverb "a picture is worth a thousand words".
The computer graphics domain “Non-Photorealistic-Rendering (NPR)” deals,
among other things, with automatic generation of scientific
illustrations and artistic drawings from the same data, photorealistic
images can be produced of. In the process, artistic painting techniques
and knowledge of illustrators are mimicked or emulated.
Those NPR-techniques have been successfully applied to the visualization
of volumetric data (e.g. CT-data in medicine). Edge enhancement,
drawing of contour lines and dyeing with pseudocolor are only a
few examples of methods for improving navigation in data and
accelerating the location of interesting information.
Because they create such complex views of data, a significant part of
the numerous existing techniques in information visualization
are very difficult to understand without intensive instruction or
interesting information is hard to see. Therefore,
in order to improve the communication of information
from automatic generated representations, it is the aim of my work
to transfer methods and techniques from illustration / NPR to the
field of information visualization thus easing the access to
complex visualization.
|
