Options
1991
Conference Paper
Titel
Advanced visualization systems. From parallel rendering to neural network imaging
Abstract
The following paper gives an introduction to advanced computer graphics systems and describes a new approach to remote sensing data classification using artificial neural network models. First of all, scientific visualization as a new technique to present large data sets is explained as well as rendering techniques for photorealistic image generation. Furthermore, advanced graphics hardware architectures and present user interface toolkits, as the highly interactive X-window system are described. The second part of the paper is dedicated to a new approach to remote sensing data classification using neural networks, which are able to include texture and structure and to cluster arbitrary density functions in feature space. Basic paradigms as back propagation or the Kohonen map are introduced and first results of the software are presented.