Research of Crop Canopy Structural Parameter by Using Hyperspectral Vegetation Indices of Cotton
|Course||Crop Cultivation and Farming System|
|Keywords||Cotton Hyperspectral Hyperspectral vegetation index Canopy structure parameters Remote Sensing Model|
The article uses the American the ASD FieldSpecProFR back hanging Spectroradiometer field measured at different growth stages of cotton canopy hyperspectral data, get the cotton canopy structure over the same period. EXCEL and other data analysis tools and data analysis, statistical software, through methods such as multivariate statistical techniques to investigate the relationship between hyperspectral vegetation indices and corresponding canopy structure and the establishment of the estimation model and accuracy test indicators evaluated, the final Screening to determine optimal hyperspectral estimation model to estimate the structural parameters of the canopy. By analyzing the correlation between cotton canopy structure parameters, hyperspectral vegetation indices and both show that: the high spectral vegetation index RVI, NDVI, PVI, DVI, RDVI, PRI, VARI-700 MGVI, NDI and SAVI and AFM reached a very significant with correlation, to RVI reached 0.8279; vegetation index RVI, NDVI, VARI-700, MNSI, and MSAVI with LAI also reached 1% significant level, to RVI's largest, reached 0.8353; vegetation index RVI, NDVI, VARI-700,, the of GVI, MYVI ASBI, AGVI and NDI and ADM correlation RVI largest, reaching 0.7164. Visible use of RVI better estimate the feasibility of AFM, LAI and ADM. Which can be constructed canopy structure parameters inversion model hyperspectral vegetation indices as independent variables, the use of remote sensing technology for the production of large-area, non-destructive, real fast to monitor the growth of cotton provided major technical basis. Study cotton canopy hyperspectral vegetation index with the variation of the crop growth period were cotton spectral data and canopy cover regression analysis. 1%: the ratio vegetation index (RVI) and the linear correlation of the canopy cover of the very significant level (r = .6735 **, n = 32), and can take advantage of the RVI inversion cotton canopy coverage; established vegetation index the correlation coefficient of the quadratic function model the highest (r = .7161 **, n = 32), total root mean square error RMSE 0.1527g/m2, can be used to extract cotton canopy coverage for production using The remote sensing timely provide an important basis for evaluation cotton growth situation. 3, on the different growth stages of cotton canopy structure parameters (MFIA, TCDP, TCRP, K, MLD, LAI, AFM and ADM) with 32 hyperspectral vegetation index correlated statistical analysis, the cotton canopy structure parameters and the correlation coefficients of the 32 high spectral vegetation index, to find out the best correlation of hyperspectral vegetation index, to lay the foundation for modeling using hyperspectral vegetation index. Based on the relationship of cotton each canopy structure parameters and hyperspectral vegetation indices, to identify the best spectral vegetation index, and the establishment of relevant statistical model based on hyperspectral vegetation index of cotton canopy structure parameters.