Hyperspectral Unmixing Based on Nonnegative Matrix Factorization: A Comprehensive Review
This is an incremental survey for researchers in remote sensing and image processing, summarizing existing techniques without introducing new methods.
The paper provides a comprehensive review of nonnegative matrix factorization (NMF) methods for hyperspectral unmixing, categorizing developments and illustrating algorithm effectiveness through experiments.
Hyperspectral unmixing has been an important technique that estimates a set of endmembers and their corresponding abundances from a hyperspectral image (HSI). Nonnegative matrix factorization (NMF) plays an increasingly significant role in solving this problem. In this article, we present a comprehensive survey of the NMF-based methods proposed for hyperspectral unmixing. Taking the NMF model as a baseline, we show how to improve NMF by utilizing the main properties of HSIs (e.g., spectral, spatial, and structural information). We categorize three important development directions including constrained NMF, structured NMF, and generalized NMF. Furthermore, several experiments are conducted to illustrate the effectiveness of associated algorithms. Finally, we conclude the article with possible future directions with the purposes of providing guidelines and inspiration to promote the development of hyperspectral unmixing.