Nitrogen contents of rice panicle and paddy by hyperspectral remote sensing.

The nitrogen content or crude protein content in rice grains is one of the important indices to evaluate the nutrition and taste quality of rice. Normal determination of their contents by chemical methods is highly expensive and time consuming. The hyperspectral reflectances of the canopy, flag leaf and panicle of 5 rice varieties are measured by a ASD FieldSpec Pro FR in field under 3 nitrogen support levels in maturing process. The nitrogen contents of stems, leaves, flag leaves, panicles and rice paddy and their crude protein contents are determined. The correlation among them is analyzed. The panicles nitrogen contents (%) are very significantly correlate not only to that of stems, leaves and flag leaves and chlorophyll contents (mg g(-1)) of flag leaves at milking and maturing stages, but also to the spectral reflectance rho(lambda), the first derivative spectra D(lambda) and RVI of canopy, flag leaf and panicle itself. The nitrogen contents (%) of rice paddy are very significantly correlative to that of stems and leaves and the spectral reflectance rho(lambda), the first derivative spectra D(lambda) and RVI of canopy at some wave bands at booting, milking and maturing stages. For the squared multiple correlation coefficients (R2) of estimating the nitrogen contents of panicle and paddy by canopies spectra, we find R2 > 0.80 at milking stage, R2 > 0.75 at maturing stage, but for the estimation of panicle by the spectra of flag leaf and panicle itself, we have R2 > 0.65. It indicates that it can be feasible for estimating the contents of nitrogen and crude protein in rice grains by hyperspectral remote sensing. It provide basis for monitoring rice quality by remote sensing.