Chlorophyll And Leaf Area Indexestimation Based On Hyperspectrum Of Summer Corn

It is one of the key technologies to solving the problems and the basis for the implementation of precision agriculture to get the information of crops on the ecological environment of farmland accurately and simultaneously. In this paper, we take the summer corn as the research object, after learning the hyperspectral remote sensing, mathematical statics and other technical methods, analyzied the quantitative relationship in the growth process between the spectral reflectance and the temporal variation of the chlorophyll content(SPAD) and leaf area index(LAI) of the summer corn, constructed estimation model of canopy spectral at different growth stages based on the original spectral reflectance characteristic parameter, ‘trilateral’ parameter based on first derivate spectral reflectance, realized the canopy hyperspectral monitor of chlorophyll content and leaf area index of the corn, tested and verified the accuracy of the estimation model by using evaluation indexes including the coefficient of determination(R2), root mean square error(RMSE) and the relative error(RE), so then guide the field management of the corn. The main conclusions of the papers are as follows:(1) The curve of the summer corn canopy spectral reflectance match the curve of the green health plant, there are significant differences in different growth stages. As time goes the summer corn growth stage, the reflectance of canopy spectral has increased gradually in the visible light, after the increase then it reduced in the near infrared band, the reflectance reach the peak in the silking stage. The red edge parameters of the summer corn(680-760 nm) are very different in different growth stages, the position of red edge(λr) moves forward to the longer wavelength from the seven leaves stage to the silking stage, otherwise, it moves to the shorter wavelength from the silking stage to the ripe stage, the tendency of the red edge slope(Dr) and red edge area are consistent with the position of the red edge, first and then decrease after reaching the peak in the silking stage.(2) There are obvious differences between the original reflectance and red edge parameters of the summer corn under the condition of different chlorophyll content(SPAD), the value of SPAD lager, the reflectance of canopy spectral is smaller in the visible light and greater in the near- infrared band; When the position of red edge moves to the longer wavelength, the slope and the area of red edge are lager. In the model to invert the chlorophyll content(SPAD) of the summer corn through the original parameters( "Red Valley" "Green Peak"), the first derivative spectra "trilateral" parameter( "blue edge", "yellow edge" and "red edge") of the characteristics of canopy spectral, the model based on the "trilateral" parameter more accurate than the model based on the original spectral reflectance comparing the value of the coefficient of determination(R2), root mean square error(RMSE) and relative error(RE). When using the original reflectance of the canopy spectral to invert the chlorophyll content, the priorities of seven leaves stages are the reflectance and the area of the green peak, the accuracy of the parameters of the green peak are higher than the red valley during the silking road and the milk stage, the reflectance of the green peak and the area of red edge can be chosen to fitting in the ripe stage. When using the ‘trilateral’ parameters to invert the value of SPAD, the fitting effect are the best when using the parameter of the yellow edge in the seven leaves stage, the inversion precision of the blue side reflectance, the blue side of the area, normalized difference vegetation index are highest in the silking stage, meanwhile, the position of red edge and the area of the blue edge are preferred to invert the value of SPAD both in the milk stage and the mature stage. Overall, the fitting accuracy of the canopy reflectance and the chlorophyll content of the summer corn in the silking stage and milk stage.(3) It shows that there are differences between the original reflectance and red edge parameters of the summer corn under the condition of different leaf area index(LAI), with the increase of the LAI, the reflectance of canopy spectral is smaller in the visible light and greater in the near- infrared band; When the position of red edge moves to the longer wavelength, the slope and the area of red edge are lager. In the model to invert the chlorophyll content(SPAD) of the summer corn through the original parameters( "Red Valley" "Green Peak"), the first derivative spectra "trilateral" parameter( "blue edge", "yellow edge" and "red edge") of the characteristics of canopy spectral, the model based on the "trilateral" parameter more accurate than the model based on the original spectral reflectance comparing the value of the coefficient of determination(R2), root mean square error(RMSE) and relative error(RE). When applying the "Red Valley" "Green Peak" characteristic parameters to invert the leaf area index of the summer corn, the Seven leaves stage, the silking stage, the milk stage and the reflectance and area of green peak, and the normalized difference vegetation index had better fitting accuracy, the inversion effect is poor in the ripe stage. When using the ‘trilateral’ parameters to invert the LAI, the fitting effect are the best when using the parameter of the yellow edge in the seven leaves stage, the inversion precision of the blue side reflectance, the blue side of the area, normalized difference vegetation index are highest in the silking stage, meanwhile, the position of red edge and the area of the blue edge are preferred to invert the value of SPAD both in the milk stage and the mature stage.