Nitrogen Diagnosis And Growth Monitoring Of Winter Wheat Based On Visible And Near-infrared Spectrum Technology

Spectral characteristics of different nitrogen levels of winter wheat were researched and correlation between spectral characteristic from H-spectrum, A-spectrum, I-spectrum and nitrogen, growth parameters was analyzed.On the base of measured value analysis and validation, optimal spectral characteristic parameters and growth stage were put forward, and the predicted models were built for winter wheat nitrogen diagnosis and growth monitor. The main conclusions from the study were as followed:The difference between various waveband positions of winter wheat canopy H-spectrum and its transform affected from nitrogen treatment was existed. Diversity of correlation between spectra characteristics and nitrogen content and growth status was obvious. Great significant correlation between R550, R680 and leaf nitrogen content was found. The precision of nitrogen predicted model was high based on the reflectance of the two wavelengths. Red and blue shift showed with the nitrogen application increasing and growth stage process. The red edge parameters have the great significant relationship with leaf nitrogen content. The nitrogen predicted model built based on red edge parameters may be regarded as the optimal predicted model of nitrogen after evaluated by measured value. And the REP can estimate leaf chlorophyll content, above ground biomass and yield of winter wheat. Great significant correlation was found between Area672, ND672 and leaf nitrogen content. Based on the above two absorption parameters, leaf chlorophyll content can be calculated by linear mathematics form. The plant leaf fresh weight of winter wheat can be estimated by using GreenNDVI parameter. Correlation of H-spectrum characteristics parameters and leaf nitrogen in jointing stage was higher than other growth stages. Jointing stage was concluded as the best growth stage for H-spectrum nitrogen diagnosis. Linear regression mathematics was the main form which described the relation of spectral characteristics and nitrogen and growth status.Nitrogen application had great effect on leaf color index of different color space and its transforms, especially on R, G, I, S, M, I1, I2 and I3 of the RGB, HIS and OHTA color space. Nitrogen content can be predicted by the HIS space. Significant correlation was found between characteristic indices of the five color space (RGB,HIS,OHTA,CMYK,Lab) and leaf chlorophyll content of winter wheat. Correlation coefficient of OHTA space and nitrogen was by far the highest. Above ground biomass and plant leaf fresh weight had the significant relationship with most of color indices. Linear and exponential forms were the main forms which could describe the leaf nitrogen content. Above ground biomass has good relationship with CMYK space and RGB space.SPAD reading from SPAD-502, NDVI value from GreenSeeker, leaf nitrogen content and growth status were closely related with growth process of winter wheat. Jointing stage was the best period for leaf chlorophyll content predicted by SPAD-502 and nitrogen diagnosis by GreenSeeker. Heading stage was much more difference of growth parameters of various nitrogen treatments than other growth stage. SPAD-502 could determine leaf chlorophyll content. NDVI from GreenSeeker can diagnosis nitrogen status. Growth status can be monitored by using SPAD reading and NDVI value in heading stage.Nitrogen diagnosis using H-spectrum data was feasible. Many spectral parameters can be chosen and the precision of the predicted model was high. Compared with canopy H-spectrum, SPAD and NDVI had stronger single pertinence for crop. It was better for measuring leaf chlorophyll content using SPAD and for diagnosing nitrogen using NDVI than using H-spectrum data. Jointing stage was better for H-spectrum data measurement. Compared with it, SPAD, NDVI was better for nitrogen diagnosis in jointing and growth monitoring in heading. On the whole, SPAD was more suitable for diagnosing individual parameter of crop, and NDVI was more suitable for reflecting the population index. Compared with H-spectrum and A-spectrum, the I-spectrum can estimate leaf chlorophyll content and leaf nitrogen content, but growth status monitoring by I-spectrum need further study in future.