Research Of Crop Canopy Characteristic Information By Using Hyperspectral Remote Sensing Data

Hyperspectral resolution ratio remote sensing (Hyperspectral Remote Sensing) mean that utilizing a lot of very narrow electromagnetic wave bands to obtain spectral data from goal concerned, hyperspectral based on spectroscopy. Hyperspectral remote sensing has resolution ratio of wave band width less than 10 nm ,the wave bands are continuous (there are hundreds of wave bands within the range of 400-2500 nm ), It contains a great of information within hyperspectral data. Hyperspectral remote sensing device can describe the hyperspectral characteristics of goals, and meanwhile it can convert hyperspectral data into an image (imaging hyperspectral remote sensing), so, it can not merely be used for the classification of the crop and vegetation distinguishing, it also be used for monitoring growing status of the crops, and deriving the physical and chemical variable of the crops. The green plant has its unique reflectance characteristics, and this typical spectral characteristic is also possessed by crop. Research on the relationship between the biochemical, biophysical variable and hyperspectral remote sensing data has achieved much improved indeed. It shows a huge potential capacity for hyperspectral in the many fields, such as in deriving crop biophysical parameters and biochemical concentration, crop canopy information obtained, crop yield-estimating, plant diseases forecasting, and others'else. This essay focus on cotton as research objects by using hyperspectral technique. The hyperspectral non-imaging spectrometer ASD Field Spec Pro VNIR 2500 was used to record cotton hyperspecctral data at five key growing stages in Xinjiang. Analysis of reflectance and derivative spectral data characteristics, and utilizing Spectral Derivative, Multivariate Regression Modeling and Spectral Position analysis methods analyze relationship between cotton canopy parameters and hyperspectal data. Actualize cotton canopy information with quick, valid, non-touch, and non-destruction field data collection and processing, building quantitative deriving modeling, in order to provide an approach of analysis, simulation, evaluation, and design cotton canopy characteristic parameters. The main biological physical parameter includes: leaf area index (LAI), K, MFIA, PAR, Fresh biomass and dry biomass, primary hyperspectral data treatment uses EXCEL. The results indicated that Based on NDVI estimate cotton LAI, the model optimum y=11.084x12.024,R=0.8076**; based on RVI estimate cotton fresh biomass, the exponential function model optimum y=52.261.exp(0.1024x),R= 0.8114**; based on RVI estimate cotton dry biomass, exponential function model optimum y=9.5552.exp(0.1133x),R= 0.8330**.The essay would provide quantitative target for monitoring field cotton growth and actualizing real time diagnosis on cotton, enhancing cotton field management level, and improving regulation effect of cotton canopy, provide evidences. Finally, this is very significant for whole crop field management level enhancement in Xinjiang.