Band Image Analysis from Oil Spill Hyperspectral Data

On oil spill surveillance, remote sensing is an important platform for monitoring, The Pushbroom Hyperspectral Imager(PHI) is to obtain high resolution hyperspectral images of important means. Taking the PHI acquired oil spill hyperspectral data as an example, analyzes the features of image information of hyperspectral data, considering the information content of each band, band correlation and spectral absorption characteristics and separability and other factors on the basis of oil spill water, normal sea water two the difference of spectral curve analysis, According to the oil spill water, normal seawater difference maximum, extraction of single band image ( wavelength of 581.8nm ); the use of principal component analysis oil spill hyperspectral data, will have a correlation of 124 bands hyperspectral image compression to completely separate the three remote sensing image, a clear access to oil spill details; using ISODATA algorithm (based on the minimal spectral distance formula),unsupervised classification of oil spill and seawater analysis.