Nitrogen Stress Characteristics Of Pepper Leaves Based On Spectral Technology

Nitrogen is essential to plant growth and is the basic constituent element of protein,chlorophyll and other key organic molecules.Nitrogen deficiency will cause problems such as weak plants,slow growth and leaf loss of green,while excessive application of nitrogen fertilizer is easy to cause soil salinization and damage the environment.Therefore,aiming at the application requirements of rapid identification of crop nitrogen stress in agricultural precision fertilization and variable fertilization,this paper took multi-variety growing period of pepper as the research object to analyze the changes of nitrogen content and chlorophyll content in leaves of different varieties of pepper and nitrogen stress time.The selection method of chlorophyll fluorescence parameters related to nitrogen stress was studied,and the characteristic parameters of chlorophyll fluorescence highly correlated with nitrogen stress were extracted to explore the classification method of pepper nitrogen stress state based on chlorophyll fluorescence parameters.Taking the classification of nitrogen stress states as samples,the spectral characteristics were analyzed,the spectral characteristic wavelength was extracted,and the discrimination model of pepper nitrogen stress state based on spectral information was constructed,in order to realize the early identification of pepper nitrogen stress and judge the nitrogen stress state,and to provide theoretical basis and technical guidance for the precise application of nitrogen fertilizer combined with crop growth requirements.The main research contents and conclusions of this paper are as follows:(1)Design of pepper nitrogen stress test.The growth period of 6 varieties of pepper was taken as the research object,and the hydroponics method was used to accurately control the nitrogen nutrient supply conditions.The sampling period of each parameter was determined to be 4 days through the pre-experiment.Leaf reflection spectrum data,leaf chlorophyll fluorescence parameter data,leaf total nitrogen content data and leaf chlorophyll content data were collected at the 1st,5th,9th,13 th,17th and 21 st days of nitrogen stress,respectively.Among the spectral data,the band range is 400-900 nm,and there are 24 types of chlorophyll fluorescence parameters,which provide data support for future research.(2)Effect of nitrogen stress on pepper growth parameters.The changes of physiological parameters such as nitrogen content,chlorophyll content,plant height,stem diameter and chlorophyll fluorescence parameters were analyzed under nitrogen stress.The results showed that,with the increase of nitrogen stress time,the nitrogen content of pepper leaves showed a decreasing trend,which was the lowest to 24.91mg/g.Except for B35,the nitrogen content of other 5 varieties decreased obviously.The chlorophyll content of pepper leaves of all varieties showed a trend of fluctuation decrease.The Fv/Fm parameters of chlorophyll fluorescence decreased significantly in the first 13 days,and decreased to0.66 after 13 days and fluctuated in a small range.Compared with the control group,the growth rate of plant height,stem diameter and other growth parameters of the experimental group decreased significantly.After comprehensive analysis,considering individual differences and differences among varieties,it is concluded that considering the subsequent discrimination of nitrogen stress degree of pepper,using a single index classification may cause errors.(3)Study on classification method of nitrogen stress of capsicum L.In order to solve the problem that the time sequence variation rules of various parameters of pepper under nitrogen stress vary greatly among varieties,the effects of nitrogen stress on growth parameters of pepper were analyzed based on chlorophyll fluorescence parameters of pepper leaves,and the classification method of nitrogen stress degree based on plant physiology/growth state was explored.In this paper,Pearson correlation analysis method was used to find out the chlorophyll fluorescence parameters associated with strong nitrogen stress in varieties,and the relevant parameters among varieties were fused.Finally,four groups of fluorescence parameters such as Fo,Fm/Fo,Fv/Fo and Fv/Fm were determined as the basis for classification.All nitrogen stresses were divided into three categories: normal,mild and severe.The K-means clustering algorithm was used to perform unsupervised clustering on the nitrogen stress data of various varieties of capsaicum.The contour coefficient method was used for internal analysis of the results,and the chlorophyll content and nitrogen content were used for external analysis of the clustering results.(4)Identification model of pepper nitrogen stress based on spectrum.The spectral data were used to construct the discriminant model,and the classification results obtained by the classification algorithm in the previous chapter were verified.Firstly,abnormal data screening was carried out on the obtained chlorophyll fluorescence parameters and leaf reflection spectral data,and then the leaf spectral data was preprocessed by using SNV,MSC,S-G filtering and other methods.Then the feature data of spectral band data was extracted according to the classification results obtained by K-means clustering algorithm.SPA and CARS were used for feature extraction.After feature extraction,SVM algorithm and BP algorithm were used to construct the discriminant model.The final test results showed that,among the discriminant degree models,S-G pretreatment method and CARS feature extraction method had the highest discriminant accuracy,with the overall recognition rate of 95%,and the discriminant rates of 98%,94% and 89% for normal group,mild stress and severe stress,respectively.