A new method on class-of-interest oriented spectral unmixing based on nonnegative matrix factorization

摘要

Spectral unmixing is a challenging task in hyperspectral relative problem. Nonnegative matrix factorization (NMF) is one of methods proposed to deal with hyperspectral data. But the existing methods are presented to estimate proportions of all end members in the hyperspectral data. However, we need to estimate proportions of class-of-interest rather than all constituent materials. If we only estimate class-of-interest abundance information, then other materials will be regarded as interference existing in the data. In this paper, a new method is proposed that based on least squares algorithm estimated for abundance and minimum distance constrained nonnegative matrix factorization (LSMDCNMF). This method can estimate the class-of-interest abundance information with the consideration of non-class-of-interest. Experimental results show that the proposed method can produce competitive results than other methods.