Oil spill detection using marine SAR images

A probabilistic approach to distinguish oil spills from other similar oceanic features in marine Synthetic Aperture Radar ( SAR) images has been developed and tested. The method uses statistical information obtained from previous mesurements of physical and geometrical characteristics for both oil spill and natural features. A sample image is evaluated using two different procedures to determine the probability that it is an oil spill, the results of the two procedures are then compared. The classification-algorithm performance was evaluated using a test dataset containing 80 examples that were oil spills and 43 that were natural features exhibiting characteristics similar to oil spills: more than 80% of the samples were classified correctly. The reliability of the method was then determined using a new dataset and similar results were obtained.