A neural network architecture and training procedure that are suitable for texture segmentation and labeling are developed. The underlying premise of the approach is that texture segmentation can be achieved by recognizing local differences in texels. The proposed architecture comprises a feature extraction network and a texture discrimination network, which is really a variation of T. Kohonen's adaptive learning network (Self-Organization and Associative Memory, Springer-Verlag, 1984). Results from a preliminary computer simulation are presented.
A neural network architecture for texture segmentation and labelling
Analoge Möglichkeiten des BSB-Modells am Beispiel die nicht-strahlenden Umkehr-Flugkörper Problems
IJCNN: International Joint Conference on Neural Networks ; 1 ; 127-133
1989
7 Seiten, 14 Quellen
Aufsatz (Konferenz)
Englisch
Texture and Neural Network for Road Segmentation
British Library Conference Proceedings | 1995
|Road Texture Segmentation with a Neural Network
British Library Conference Proceedings | 1997
|