PROJECT TITLE :

A Multilinear Mixing Model for Nonlinear Spectral Unmixing

ABSTRACT:

In hyperspectral unmixing, bilinear and linear–quadratic models have become in style recently, and additionally the polynomial postnonlinear model shows promising results. These models do not take into account endmember interactions involving a lot of than two endmembers, though such interactions would possibly compose a nontrivial half of the observed spectrum in scenarios involving bright materials and advanced geometrical structures, such as vegetation and intimate mixtures. In this paper, we present an extension of these models to include an infinite variety of interactions. Several technical problems, like divergence of the resulting series, will be avoided by introducing an optical interaction likelihood, which becomes the only free parameter of the model in addition to the abundances. We tend to gift an unmixing strategy based mostly on this multilinear mixing (MLM) model; gift comparisons with the bilinear models and therefore the Hapke model for intimate mixing; and show that, in many scenarios, the MLM model obtains superior results.