Detecting And Mapping Kaolinite In The Algerian Central Hoggar With A Partial Linear Nmf-Based Unmixing Method

In this work, a remote sensing unmixing-based method is used for detecting and mapping kaolinite deposits in the Algerian Central Hoggar. The used approach, related to linear spectral unmixing techniques, is based on a partial multiplicative nonnegative matrix factorization algorithm. This one aims to unmix the considered remote sensing data by exploiting some known kaolinite spectra. Experiments are conducted on real Advanced Spaceborne Thermal Emission and Reflection (ASTER) data that cover the considered study area to evaluate the potential of the used approach. The obtained results are mainly confirmed and validated by field investigations.