Feature extraction and classification of ocean oil spill based on SAR image

The detection of ocean oil spill based on synthetic aperture radar (SAR) image has been a hot topic attracting extensive attention. In this paper, a hybrid scheme, in which we extract feature parameters and then achieve classification as follows, is presented. Two-dimensional (2-D) Otsu algorithm is applied in image segmentation process, and neural network is applied in classification course. Before image segmentation, a sort of universal processing is used, and it enables 2-D Otsu algorithm to be more applicable to SAR images of ocean oil spill.