Hyperspectral remote sensing estimation model for rape moisture content under different levels of water stress

It was developed that using spectral analysis to quantitatively analyze the rape water stress condition. Region stepwise regression was proposed to select the characteristic wavelength of rape moisture content and to establish the model between the moisture content and the spectral reflectance at characteristic wavelength of rape. From the spectral datum of rape samples under different water levels, it was found that the linear relationships between moisture content of rape and spectral reflectance value at 960nm, 1450nm and 1650nm are very notable. The correlation coefficient between the predict value and the measured value is 0.87, the RMSE is 4.57. The results show that the method of stepwise regression could select the characteristic wavelengths of rape moisture content accurately and enhance model forecast precision.