The Processing Of Hyperion Data And The Application Of Hyperspectral Remote Sensing Technology In Karst-Desertification Classification

After having undergone the stages of Pan, Colour Photography, Multispectral Remote Sensing, remote sensing technology is heading towards a brand-new era - Hyperspectral Remote Sensing. Hyperspectral Remote Sensing data have been proved its superiority of providing highly desired information together with more contiguous and narrower spectral bands, as it overcomes the distortion in accuracy and scope. On one hand, desertification is among the most challenging environmental and social problem worldwide in the current scope. Meanwhile, in the region of karst in the south of China, it appears as Karst-Desertification. This paper focuses on the reflection of Karst-Desertification in the area of remote sensing by means of the Hyperspectral Method and the classification method in the light of the sensitive correspondence of Hyperion to Karst-Desertification.As one of the main data sources, Hyperion data, scanned by Earth Observing-1 (EO-1) satellite, are employed by the fieldwork to investigate the situation of Karst-Desertification on the border between Anshun and Qianxinan. With the help of the fieldwork, we gained a data spectral curve for the research region, and built up a single-surface material spectral library. In order to acquire reflectance information, the essential pre-processing upon Hyperion data was carried out. The karst-desertification classification spectral library, based on Hyperion data, was established upon 3S technology that integrates Hyperion data, topographic map, lithology and petrofabric map, slope map, investigation route, GPS point, spectral curve and photo, both synthetically and theoretically. Following the presentation of the fieldwork results, the paper discusses five prevailing classification methods and steps for pre-processing, providing some ideas for following work.Hyperspectral Remote Sensing data based karst-desertification classification spectral library is more significant in this paper is highlighted in this paper. In addition to the introduction on the application of Hyperspectral Remote Sensing data in karst region, the improvement of karst-desertification characteristic standard, based on traditional Multispectral Remote Sensing data, is also performed in our research. Furthermore, the new karst-desertification characteristic standard, rooted in the spectral curve from Hyperspectral Remote Sensing images. The accuracy and practicability arise from RS interpretation have been ulteriorly improved.