Comparison and evaluation between several spectral based endmember extraction algorithms on real hyperspectral remotely sensed images

Mixing constituent elements at the pixel level occurs as the result of insufficient spatial resolution in hyperspectral images. Extraction of spectral signatures of such constituent elements which are known as endmembers in each mixed pixel and estimating the abundance maps of such pure spectral signatures are two important roles in Spectral Mixture Analysis(SMA). Most of algorithms, which was proposed for endmember extraction, are relayed on only spectral information such as OSP,N-Finder,VCA,PPI,IEA, SGA without considering spatial arrangement and distribution of image pixels. We review these algorithms and compare them from the view points of RMSE and processing time, on two real hyperspectral Aviris Cuprite and Indian Pines scenes. We demonstrated that VCA results the best response to RMSE of reconstructed original image and processing time in comparison with others.

کلیدواژه ها:

endmember extraction ، spectral ، unmixing ، hyperspectral image