Factors Analysis About Alteration Minerals Information Extraction Using Simulation HyMap Data

Compared with traditional optical remote sensing technology，hyperspectralremote sensing technology has its unique advantages that a hyperspectral remotesensing image contains a space, radiation and spectral triple information which raisesspectral dimension on the base of the traditional two-dimensional image sensing.Hyperspectral remote sensing technology plays an important role in regionalgeological mapping and mineral exploration.In general, A hyperspectral remote sensing images contains dozens or evenhundreds bands. Because of huge amount of bands, assessing the quality of the data isa complex task.. According to characteristics of correlation in the bands, I useK-means clustering analysis to reduce the data from hundreds of bands to eightbands,then use the methods of histogram, to-noise ratio, modulation,information entropy to analysis quality of the data.Diagnostic spectral characteristic is the basis for identification of differentminerals. In the grasp of the spectral characteristics of minerals, I will analyse thefactors about the spatial resolution, spectral resolution, band selection and mineralmapping method when we extract information from simulated HyMap Data.In this paper, I will extract mineral components combining Isodata classificationand PPI. Basic ideas: first I use the methods of isodata for hyperspectral imageclassification method; then, I calculate pure index of category in classification results;the last, I choose mineral components according to the value of pure index.On the basis of the above analysis, I extract alteration information in the studiedarea and obtain good results.