Slick detection in SAR images

A study of oil slicks and their look-alikes in SAR imagery is performed. Slicks are classified according to characteristic image expressions, backscatter profiles and gradients, geographical occurrence and weather limitations. A conceptual model for distinguishing oil slicks from other slick expressions is proposed. The model perform best when auxiliary data is available. The radar backscatter cross section and the radar contrast or damping ratio, defined as the ratio of the backscattered radar power from slick free and slick covered surfaces, are calculated for the different types of slick surfaces. Finally, a comparison between experimental results from a laboratory study and the SAR data is performed.<<ETX>>