Analysis of hyper-spectral remote sensing features of oil film on nearshore waters

The reflectance spectrum and hyper-spectral images were obtained via a portable spectroradiometer and an imaging spectrometer respectively.Some typical features of oil film reflectance spectrum were extracted after the exploration of the reflectance and their first-order derivatives.The image was classified using the end-member's spectra based Spectral Angle Mapping (SAM) method.The result showed that,for the sea water,thin oil film and thick oil film,the range of visible-near infrared was the optimum to distinguish one from the others and the range of 1000~2500nm was more suitable for the thick oil film to be identified.