Abstract
A supervised classification method using a self-organizing map (SOM) is proposed to classify remote sensing data. SOM has a characteristic that a probability density function of input data is represented as a feature map. The proposed method is realized by creating a category map from the feature map of SOM. The category map can visualize characteristics of SPOT HRV data and it is also employed as a supervised classification method. The proposed method extracts liquefied area in Kobe (Japan) damaged by the 1995 Hyogoken Nanbu earthquake using the SPOT HRV data and the category map. As an experimental result, it is shown that classification accuracies of the proposed method are higher than those of the maximum likelihood and the back-propagation methods.
