Abstract-Hyperspectral image unmixing is a source separation problem whose goal is to identify the signatures of the materials present in the imaged scene (called endmembers), and to estimate their proportions (called abundances) in each pixel. Usually, the contributions of each material are assumed to be perfectly represented by a single spectral signature and to add up in a linear way. However, the main two limitations of this model have been identified as nonlinear mixing phenomena and spectral variability, i.e. the intraclass variability of the materials. The former limitation has been addressed by designing non linear mixture models, while the second can be dealt with by using (usually linear) space varying models. The typical example is a linear mixing model where the sources can vary from one pixel to the other. In this letter, we show that a recent variability model can also estimate the abundances of nonlinear mixtures to some extent. We make the theoretical connection between nonlinear models and this variability model, and confirm it with experiments on nonlinearly generated synthetic datasets.
International audienceWhile Blind Source Separation (BSS) for linear mixtures has been well studied, the problem for nonlinear mixtures is still thought not to have a general solution. Each of the techniques proposed for solving BSS in nonlinear mixtures works mainly on specific models and cannot be generalized for many other realistic applications. Our approach in this paper is quite different and targets the general form of the problem. In this advance, we transform the nonlinear problem to a time-variant linear mixtures of the source derivatives. The proposed algorithm is based on separating the derivatives of the sources by a modified novel technique that has been developed and specialized for the problem, which is followed by an integral operator for reconstructing the sources. Our simulations show that this method separates the nonlinearly mixed sources with outstanding performance; however , there are still a few more steps to be taken to get to a comprehensive solution which are mentioned in the discussion
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.