In this thesis, we introduce a novel distributed version of the N-FINDR endmember extraction algorithm, which is able to exploit computer cluster resources in order to efficiently process large volumes of hyperspectral ...