Oil Spill Observation by means of Polarimetric SAR data
Maurizio Migliaccio(1) , Attilio Gambardella(2) , and Massimo Tranfaglia(3)
Università degli Studi di Napoli Parthenope,
Via Acton, 38,
(2) Università degli Studi di Cagliari, Piazza D’Armi, 19, 09123 Cagliari, Italy
(3) Università degli Studi del Sannio, Palazzo Bosco Lucarelli , 82100 Benevento, Italy
A study on sea oil spills observation by means of polarimetric Synthetic Aperture Radar (SAR) data is accomplished.
It is commonly accepted that oil spill detection procedures can be framed into three fundamental phases: dark patch detection, features extraction and, oil spill and look-alike classification. The polarimetric approach presented in this paper is based on the combined use of a polarimetric Constant False Alarm Rate (CFAR) filter to detect dark patches over SAR images and the Target Decomposition (TD) theorem to provide extra polarimetric features to the decision process. In particular, parameters of the <H/A/Alpha> Cloude-Pottier decomposition are used to examine scattering contributions from detected dark areas.
The polarimetric CFAR filter allows a better dark area detection with respect to the classical edge detection filters and, thus, a better geometrical and statistical feature extraction. The use of the three additional polarimetric features is meant to assist classical oil spill feature classification schemes.
Experiments are conducted on polarimetric SAR data acquired on October 11, 1994 during SIR-C/X-SAR mission, named SRL-2. The data were processed and calibrated at NASA/JPL.