On the Mathematical Formulation of the SAR Oil-Spill Observation Problem

A novel approach to oil-spill classification, based on the paradigm of one-class classification, is proposed. Basically, a classifier is trained using only examples of oil-spills, instead of using oil-spills and look-alikes, as in two-class approaches. In addition, as a large number of candidate features have been considered in the literature, a feature selection algorithm, to objectively select the most effective subset, is proposed. Results on two case study datasets are reported to validate the proposed approach.