SFES2D leverages advanced, unsupervised source separation techniques such as principal component analysis (for variance maximization), independent component analysis (for kurtosis and negentropy maximization), continuous wavelet transforms, RGB color processing, and k-means clustering segmentation. These state-of-the-art methods facilitate feature extraction, dimensionality reduction of hyperdimensional geoscientific datasets, and lineament extraction with Bayesian optimization.
Developed and released by:
Bahman Abbassi Postdoctoral Fellow in Mining Engineering at Institut de Recherche en Mines et en Environnement (IRME), Université du Québec en Abitibi-Témiscamingue (UQAT)
Address: IC113, 445, boul. de l’Université, Rouyn-Noranda (Québec) J9X 5E4
Contact: For the passcode to access SFES2D, please email [email protected]