Segmentation and classification of hyperspectral data using watershed

The paper presents a new segmentation and classiſcation scheme to analyze hyperspectral (HS) data. The Robust Color Morphological Gradient of the HS image is computed, and the watershed transformation is applied to the obtained gradient. After the pixel-wise Support Vector Machines clas-siſcation, the majority voting within the watershed regions is performed. Experimental results are presented on a 103-band airborne ROSIS image, of the University of Pavia, Italy. The integration of the spatial information from the watershed segmentation into the HS image classiſcation improves the classiſcation accuracies, when compared to the pixel-wise classiſcation. Index Terms — hyperspectral images, mathematical mor-phology, segmentation, classiſcation, watershed 1