Comprehensive Overview of Linear Hyperspectral Unmixing Techniques and the HySUPP Python Package

This paper provides a comprehensive overview of linear hyperspectral unmixing techniques, categorizing them into supervised, semi-supervised, and unsupervised (blind) approaches based on the prior knowledge of endmembers. It also introduces the open-source HySUPP Python package for reproducing and benchmarking various unmixing methods.