Semi-automatic Oil Spill Detection in Sentinel-1 SAR Images at Brazil's Coast

This paper reports a methodology proposed on the matter of semi-automatic detection of oil spills using Sentinel-1 Synthetic Aperture Radar (SAR) images to improve the capability of management of government agencies for oil spills incidents over the seas. This is an initiative to create a semi-automated system of detection based on free images of Sentinel-1 using an intelligent database of samples statistics of oils, confirmed by vessels or airplanes and photo-interpreted, and oceanographic features that continues improving itself over time. The results indicate that both the VV and HH polarizations of the Sentinel-1 detected probable oil-spills and look-alikes, however, the utility of the Sentinel-1 cross-polarizations (VH and HV) in oil-spill detection could not be discarded. Other parameters will be improved and included in this study, as parameters of segmentation, besides process of data mining.