Detection Of Photosynthetic Physiological Information In Maize Leaves Based On Chlorophyll Fluorescence Spectrum Analysis

In recent years,with the rapid development of computer technology,the use of spectral detection technology to monitor the physiological state and growth environment of plants has attracted more and more attention.Among them,as an important part of active detection technology,laser induced?LIF?chlorophyll fluorescence spectroscopy technology has enabled the rapid and nondestructive detection of photosynthetic physiological information of maize leaves.Using chlorophyll fluorescence spectroscopy as the main technical means to combine the three photosynthetic physiological information of maize leaves,three corresponding prediction models of corn were constructed using mathematical modeling methods.The main research contents and conclusions are as follows:?1?Screen the best spectral processing method.A variety of smoothing methods and spectral pretreatment methods were used to process the chlorophyll fluorescence spectra of maize leaves.After comparing the pre-modeled model performances and effects with the SVM pre-modeling model,it was concluded that the best spectral pretreatment methods for this experiment were Adjacent smoothing algorithm and FD preprocessing.?2?Extract the best characteristic wavelength of the fluorescence spectrum.Combining the fluorescence spectral feature information and Gaussian fitting algorithm,the characteristic wavelength of chlorophyll fluorescence spectrum was extracted.After comparison and analysis,the five optimal characteristic wavelengths for establishing the prediction model were determined,namely F₅₀₈,F₅₁₈,F₆₇₂,F₆₈₅and F₇₃₆.?3?The optimal mathematical modeling method based on chlorophyll fluorescence spectrometry was used to determine the net photosynthetic rate of maize leaves.After extracting the five optimal characteristic wavelengths of the fluorescence spectrum after the adjacent smoothing algorithm and the FD pretreatment,the four OCW-MLR,OCW-SVM,OCW-BP,and OCW-CART spectral prediction models were established to obtain the OCW-SVM.The predictability and sensitivity are the best.The correlation coefficients of the training set and verification set of the model are 0.9247 and 0.8937,respectively,and the root mean square errors are 0.0884 and0.0908,respectively.?4?Establish a prediction model of corn photosynthetic physiological information.A linear regression model of corn net photosynthetic rate Pn and light use efficiency LUE based on the blade temperature Ti correction and a full band625-820nm SVM model were established.After comparison and analysis,the linear regression method based on the blade temperature Ti correction was more suitable as To predict corn physiological information,the correlation coefficient R and root mean square error RMESC of the training set and verification set were 0.8815,0.0967 and0.8960,0.1685 respectively;a prediction model of the SPAD value of the corn leaf blade at 625-820 nm was established and compared.The SVM modeling method has the best performance after the modeling method.The verification set correlation coefficient R and the root mean square error RMESC are 0.8953 and 0.0264,respectively.This paper constructed a prediction model of photosynthetic physiological information of maize based on chlorophyll fluorescence spectroscopy,explored the inherent laws between fluorescence spectra and photosynthetic physiological information of maize leaves,and laid a theoretical foundation for realizing the precision and intelligence of maize with high quality,high yield,and increased yield.Content and results have certain reference and reference value for other food crops...