Spectral Unmixing with Hyperspectral Datasets of AVIRISNG

The AVIRIS-NG (Airborne Visible InfraRed Imaging Spectrometer - Next Generation) data set is in a cube format in which the signature of spectrum gives each pixel of the underlying materials in that image area. The motivation of unmixing is to find a collection of pure spectral constituents. Hyperspectral data is frequently used to identify the present materials within a scene. The interested materials contain vegetation, roadways and specific targets (i.e. manmade materials, minerals etc.). For various practical applications hyperspectral unmixing is used, such as monitoring, agriculture and management of natural disaster, issues related to security, and defense. Some of the unmixing techniques are used as Linear Mixing Model (LMM), Nonlinear Mixing Model (NLMM), BiLinear Mixing Model (BiLMM) and Additive Noise Linear Mixing Models (ANLMM). In this paper, we study about the basic concepts of hyperspectral unmixing and techniques.