Abstract
Oil slicks identification can have interference from other physical phenomena on the sea, such as biogenic films and low wind zones. In this study, polarimetric SAR images were analyzed and both spatial and spectral features were combined to reduce the influence of look-alikes, especially related to low wind zones. First, nine polarimetric features were extracted and examined in terms of slick free areas, internal waves, low wind zones, and oil slicks. The analysis proved that the low wind zones cannot be distinguished from the oil slicks using the spectral features derived from polarimetric SAR due to their similar surface properties. Therefore, the object-oriented classification, which involved the shape features of low wind zones, was introduced, and oil slicks were identified successfully. However, more studies should be performed over other look-alike features because of the limited availability of polarimetric SAR data. With more satellites carrying polarimetric SAR being launched, it is anticipated that an operational oil spill monitoring system can be developed.
Acknowledgment
This work is sponsored by the Specialized Research Fund for the Doctoral Program of Higher Education of China: Method for Oil Spill Monitoring by Polarimetric SAR (Project number: 20092125120007), and the Fundamental Research Funds for the Central Universities: Research on Seasonal Change of Sea Ice of the Liaodong Gulf in China (Project number: 2011QN094). The authors would like to thank the referees for the constructive comments that helped to clarify and improve this manuscript.