Parallel implementation of endmember extraction algorithms using NVidia graphical processing units | Publicación