A Supervised Classification Method Using a Three-Layered Self-Organizing Map and Counter Propagation

Abstract

A supervised classification method using a self-organizing map (SOM) is proposed for the classification of remote sensing data. We adopt a three-layered SOM network and counter propagation learning method for multi-spectral data. The proposed classification method is employed to identify liquefied area in Kobe (Japan) that was damaged by the 1995 Hyogoken Nanbu earthquake, using SPOT XS data. The proposed method provides a category map to visualized the SPOT XS data and offers higher classification accuracy than either the maximum likelihood or back-propagation methods (BP).