A Survey on Hyper-Spectral Image Unmixing Process

Nowadays, Hyperspectral Images are been extensively used. Applications are increasing but parallely present applications are also increasing their efficiency in their consequences with the different and improved algorithms. A Hyper-spectral image contains radiance values which consist of different particles, which are mainly Endmembers and Abundance values. These values can be generated by the method of extraction. Endmembers are basically objects which are part of an image or a scene such as water, oil, soil, metals, buildings etc. Abundance values are the pixel-wise proportion of Endmembers present in an image. Unmixing of Hyper-spectral image is a process which generates Endmembers and Abundance which consequences to the classification of the different materials present in the scene.