Study oil detection using spectral separation algorithm PPI and  FPPI images hyperspectral

Keshavarz, Ahmad; Kayhanpour, Hamid Reza; Hashemizadeh, Seyed Mohammad Karim

Study oil detection using spectral separation algorithm PPI and FPPI images hyperspectral

Document Type : Review

Authors

¹ faculty member (Assistant Professor) University of Persian Gulf

² Director/ Iranian Oil Pipeline and Telecommunication Company/ Khuzestan District

³ Technical Assistant /Iranian Oil Pipeline and Telecommunication Company / Khuzestan District Busher Islamic Azad University

Abstract

Oil spills is one of major marine environment challenge. The main impacts on the marine environment can be cited such as reduction in light transmission and also oxygen absorption and breakup which these prevents water plants photosynthesis. In this research, we used SpecTIR airborne sensor data and extracted and classified oils spills due DWH which happened in Gulf of Mexico at 2010. Atmospheric correction is done, because the data was in radiance level. In the next step, pure members extracted using PPI and FPPI algorithm with 1000 iterations and then they clustered 6 classes.Pure members classified in 6 classes of oil ratio, as 100%,80%,60%,40%,20% and water, rationality.Finally, FPPI had satisfactory results as it produces Infeasibility images for overtop False Alarm problem and also it doesn't need to use all of the pure members' spectral patterns for doing an accurate analysis in image.

Keywords