Font Size: a A A

A Study Of Hyperspectral Remote Sensing Data Preprocessing And Spectral Re-building In Desertification Areas

Posted on:2003-08-03Degree:MasterType:Thesis
Country:ChinaCandidate:H Q DuFull Text:PDF
GTID:2120360065460976Subject:Forest management
Abstract/Summary:PDF Full Text Request
The study is a part of the national natural fund project "spectral re-build and quantitatively retrieves of information with hyperspectral remote data in desertification area"(NO:30070604).Compared with traditional remote sensing,some new technique and method must be developed for hyperspectral remote sensing data processing and object discerning. Taken actual situation of desertified areas and self-characteristics of hyperspectral remote sensing data into consideration,a data preprocess scheme for hyperspectral remote sensing image was presented and a spectral re-building model was developed in detail on the basis of theory and practical application in this study. Analyze and discuss the scheme's feasibility and the model's accuracy and applicability,some achievements have been obtained in theory and application.The main tasks and innovative points of this study as follow:1.Based on the spectrum anlysis,statistical characteristic and SNR(signal to noise ratio) of hyperspectral remote sensing image,some optimal bands for desertification monitoring was selected.2.A series of algorithm and program for tangent correction,radiant correction have been developed,the tangent distortion and the difference between limb radiance and sub-airplane point(nadir) radiance was corrected well.3.Apply Wavelet analysis to remove the random noises of hyperspectral remote sensing image.4.Spectrum database has been built for reasearch areas,resample the spectrum of different object in spectrum database according to the centric wavelength of the select bands.5.Analyze the spectral characteristics of typical object. Process spectrum especially the vegetable spectrum with differential calculation,normalize processing and differential calculation after normalize processing,the spectral difference of the same object was removed.6.According to the theory of imaging of imaging spectrometer,through simulating the procedure of atmospheric radiation,a linear model was developed and applied in this study to re-build spectrum and got a satisfactory result. It lay a better foundation for object interpreting and quantitatively retrieves of information.
Keywords/Search Tags:Desertification, Hyperspectral remote sensing, Data preprocess, Wavelet analysis, Spectrum analysis, Spectrum re-build
PDF Full Text Request
Related items