Hyperspectral Image Segmentation with Discriminative Class Learning

This paper presents a Bayesian approach to hyperspectral image segmentation that boosts the performance of th e discriminative classifiers. This is achieved by combining class densities based on discriminative classifiers with a Multi-Level Logistic Markov-Gibs prior. This density favors neighboring labels of the same class. The adopted discriminative classifier is the Fast Sparse Multinomial Regression. The discrete optimization problem one is led to is solved efficiently via graph cut tools. The effectiveness of the proposed method is evaluated, with simulate d and real hyperspectral images, in two directions: 1) to improve the classification/segmentation performance and 2) to decrease the size of the training sets.