Publication

Abstract : Hyperspectral imaging is an area of interesting researches in the current scenario. The HSC(Hyperspectral cameras) used in this field is having high spectral resolution and low spatial resolution and by this reason, the spectra of pixels in the acquired data will appear as mixtures of spectra of various endmembers present in that area. Here spectral unmixing comes as a major process in the hyperspectral image analysis part. Spectral unmixing is a process by which user gets the number of pure reference materials called (endmembers),their spectral signatures and their corresponding abundance maps from the acquired hyperspectral data.In this paper performance analysis of three minimum volume based geometrical approaches namely MVSA(Minimum volume simplex analysis),MVES(Minimum volume enclosing simplex),and SISAL(simplex identification via split and augmented lagrangian) are done by applying them on the real hyperspectral data set AVIRIS Cuprite,taken over Nevada,U.S and the results are evaluated with reference to U.S.G.S spectral library which is available online.