Study Of Information Extraction And Classification Of Hyperspectral Remote Sensing

Hyperspectral remote sensing technology, as one trend of remote sensing's development, has been successfully used in several fields by its higher spectral resolution. Using OMIS1 aerial hyperspectral image of Ningxia Rujigou coalfield as main information data, preprocessing of OMIS1 image which included geometrical correction, noise elimination, reflection and ground surface temperature retrieval are discussed in the paper. Such information extraction and explanation as heat information of coalfield and burnt rocks information which are indication of coal fire are then studied. Based on all above, using abundant spectral information of OMIS1 image, the author studied classification method of hyperspectral images based on spectral dimension, and evaluated the precision of classification.