A comparison of feature extraction methodologies applied on hyperspectral data

In this paper we present a comparative study for features extraction from hyperspectral data where the performance given by three different unsupervised techniques is considered. Among the three, one technique is rather innovative in the field of hyperspectral data processing and is based on neural networks algorithms. The study has been carried out for a set of hyper-spectral data collected by the Airborne Hyper-spectral line-Scanner radiometer (AHS) over a test site in Northeast Germany. The results have been quantitatively evaluated and critically analyzed either in terms of their capability of representing the hyperspectral data with a reduced number of components or in terms of the accuracy obtained on the final derived product