Spectral Data Processing And It's Application In Vegetation Information Extraction

The study on spectral characteristics in surface features is an important component of modern remote sensing technology. It is also one of the main signs for discriminating the surface features. The most important characteristic of remote sensing is the unities of map and spectrum information, and the spectrum information is an important advantage in hyperspectral technology, whose application process requires ground spectra data and picture spectrum combined together. Therefore, the collection and processing of spectral data is essential, and the quality of spectral directly affects the results of the application.In this study, the ASD FieldSpec(?) Pro FR wide-band wild spectrograph was the means of spectral data acquisition, and the design of spectrum acquisition programme, the wild spectral data measurement and spectral properties of vegetation was for the target of extracting spectral information of vegetation.Enclosed the vegetation information extraction, this paper focused on the spectral characteristics of the vegetation , spectral data processing and information extraction technology .Using remote sensing software Envi to pretreatment and transform spectral data in order to highlight the nuances of spectral data to the comparative analysis.The main content of article as follows:(1)Analysis and determine the necessary data and their types and design spectrum acquisition programme in accordance with the needs of hyper-spectral vegetation study, and using ASD spectrograph spectral properties of vegetation to measure spectral data in surface features. Because of the spectral data vulnerable to atmospheric conditions, equipment performance parameters, terrain, target itself affected by factors such as spectral characteristics,. the spectral data must be removed the equipment itself and the outside world noise conditions(2) Aimed at this experimental application study , choose appropriate way from existing spectral matching processing technologies such as Binary Encoding, Spectral Angle Mapping, Spectral Absorption Index, Spectral Feature Fitting, Matched Filter and mixed-pixel analysis of the issue. (3) On selected areas of case studies, respectively extract vegetation information and remote sensing image processing spectrum by ENVI. After acquired identification map, evaluate the results.(4) Resembling the hyper-spectral data to the ASTER data's wave band scope, establishes both relation. Contrast to the traditional classification, analyzes the superiority of ground spectral data participate in the ASTER data classification and evaluate the accuracy.