ABSTRACT
Ocean pollution caused by oil spills is currently considered a signicant environmental hazard. Synthetic aperture radar (SAR) carried by Earthobserving satellites have proven to be a powerful tool for monitoring oil spills in marine waters. This chapter provides a comprehensive overview of SAR image analysis approaches to oil-spill monitoring with a focus on dark-spot detection. Dark-spot detection is a fundamental step in computerized marine oil-spill detection systems. Automated detection of dark-spots using SAR imagery is a challenging task, primarily because of the presence of multiplicative noise known as speckle. Speckle noise must be treated appropriately and adequately to make a good use of SAR imagery for dark-spot detection. The results of the method introduced in this chapter demonstrate that dark-spots (possible oil slicks) can be effectively detected in an SAR image by integrating the total variation optimization segmentation approach. Also, the proposed dark-spot detection method presents better segmentation results and takes less computing time.
