Hyperspectral Unmixing Based on Nonnegative Matrix Factorization: A Comprehensive Review

Hyperspectral unmixing has been an important technique that estimates a set of endmembers and their corresponding abundances from hyperspectral image (HSI). Nonnegative matrix factorization (NMF) plays increasingly significant role to solve this problem. In article, we present comprehensive survey the NMF-based methods proposed for unmixing. Taking NMF model as baseline, show how improve by utilizing main properties HSIs (e.g., spectral, spatial, structural information). We categorize three development directions including constrained NMF, structured generalized NMF. Furthermore, several experiments are conducted illustrate effectiveness associated algorithms. Finally, conclude paper with possible future purposes providing guidelines inspiration promote