Qualitative SAR Dark Spot Analysis for Oil Anomaly Characterization

Abstract

The characterization of the morphology of a dark spot event in a SAR image is crucial to determining the nature, fate and classification of a potential oil on water anomaly. Hence the need to adequately analyze SAR backscatter values from an amplitude power image were darks spots have been identified and extracted. The dark spots are potential areas of oil on water events and it is imperative that they are understood sufficiently for the purpose of assigning the level of confidence in the processing of oil analysis, understanding the type of oil on water event, classifying the extent of weathering impacting the detected anomaly and relating the oil anomaly to a potential source. In order to generate high quality datasets for anomaly characterization, it is imperative to transform satellite image amplitude data into power values in units of dB and represented on a logarithmic scale. This new power dataset is corrected radiometrically and requisite speckle to noise filter applied for data cleaning and noise suppression. Oil anomaly or darks spot extraction, representative of a potential oil on water is event which is the primary objective of the pre-processing is done. Gamma enhancement was applied on the dark spots in order to characterize their radiometric, textural and geometric properties. This is predominantly for estimating the confidence level of each anomaly detected, and their properties which are indicative of the type, fate, age and physical processes relating to each anomaly.