An Admm-Based Hyperspectral Unmixing Algorithm For A Modified Almm Addressing Spectral Variability

This work presents a linear mixing model, and a related algorithm, for addressing spectral variability in hyperspectral unmixing. The research focuses on a variant of the modified Augmented Linear Mixing Model (mALMM), and the development and evaluation of an unmixing algorithm based on the Alternating Direction Method of Multipliers (ADMM) concept. The new variant of the mALMM is introduced to model spectral variability, while the ADMM provides an optimization framework for unmixing hyperspectral data. The study evaluates the performance of the designed ADMM-based algorithm in comparison to a literature technique based on nonnegative matrix factorization. The experiments are conducted using synthetic hyperspectral data to assess the algorithm's effectiveness in handling spectral variability. The results demonstrate the attractiveness of the designed unmixing algorithm for hyperspectral data with spectral variability.