Study on the Extraction and Applications of Spectral Features in Hyperspectral Remote Sensing

According to the properties and computation principles, spectral features in hyperspectral remote sensing (RS) information can be grouped into three levels: spectral curve features, spectral transformation features and spectral measurement features. Spectral curve features mainly include direct spectra encoding, reflection and absorption features. Spectral transformation features include normalized difference of vegetation index (NDVI), derivate spectra and spectral computation features. Spectral measurement features include spectral angle (SA), spectral information divergence (SID), spectral distance and correlation index. Based on the analysis to those basic algorithms, several problems about feature extraction, matching and application were discussed.