Inversion Of Chlorophyll And Nitrogen Content Distribution In Apple Leaves Based On Hyperspectral Imaging Technology

Chlorophyll and nitrogen play an important role in plant metabolism and growth.Chlorophyll in apple leaves plays an important role in photosynthesis.Nitrogen is an essential element in apple growth.Both of them directly affect the growth and yield of apple.The estimation and inversion of chlorophyll and nitrogen content in apple leaves by hyperspectral imaging technology is of great significance to the nutrient distribution and Precise Fertilization of apple leaves.The traditional methods for the determination of plant nutrients are chemical analysis,which are time-consuming,laborious and highly technical.The recent development of hyperspectral technology has the advantages of fast,convenient,low cost and non-destructive.Hyperspectral imaging technology combines the advantages of spectrum and image.It has the characteristics of high resolution,multi-band and Atlas integration.It can detect the external characteristics and internal components of objects.It can not only quantitatively analyze the nutrient content of plants by multi-band spectrum,but also visualize the spatial distribution differences of plants by image.Therefore,the use of hyperspectral imaging technology to obtain changes in leaf nutrition during apple growth can provide technical support for Precise Fertilization management.The research area is Qixia City,Yantai,Shandong Province,known as the"City of Apple in the World".The study sample was Red Fuji apple leaves in September 2016.The hyperspectral data of apple leaves were measured by imaging spectrometer.The chlorophyll and nitrogen contents of apple leaves were determined by spectrophotometer and Kjeldahl nitrogen meter respectively.The spectral information of apple leaves was extracted and pretreated.(1)The sensitive wavelengths of chlorophyll content and nitrogen content in apple leaves were screened out and the vegetation index was constructed.By multi-scattering correction pretreatment of the spectrum,the correlation between chlorophyll content and spectral data of apple leaves was improved,and six sensitive wavelengths of 712.5 nm,509.95 nm,561.222nm,840.62 nm,696.67 nm and 987.91 nm were screened.The SG smoothing first-order differential is applied to the spectrum,and the correlation between nitrogen content and spectral data of apple leaves is improved.Five sensitive wavelengths of 535.53 nm,576.69nm,633.77 nm,654.66 nm and 701.94 nm are screened to establish the model.A good model of six vegetation indices NIR/NIR(740,780),RVI2(765,787),ZTM(710,750),VLOPT2(730,760),GRVI(550,800),GNDVI(550,800)was established and compared.(2)Established chlorophyll content estimation models of apple leaves based on hyperspectral imaging technology,which were partial least squares,principal component regression and quadratic polynomial stepwise regression models.Established stepwise regression analysis model,BP artificial neural network model and ridge regression model of nitrogen content of apple leaves based on hyperspectral imaging technology,and selected the best prediction through comparative analysis.Measuring model.The results showed that the established MSC-PCA model could better estimate chlorophyll content and SG-D-SRA model could better estimate nitrogen content(chlorophyll:R~2=0.8004,RMSE=0.3188,RE=26.4%;nitrogen:R~2=0.6866,RMSE=0.3188,RE=10.21%).By using chlorophyll content to modify the nitrogen estimation model,it was found that the R~2 of the revised nitrogen estimation model prediction set increased from 0.7431 to 0.7866,and the R~2 of the test set increased from 0.6866 to 0.7071,with good results.(3)Inversion maps of chlorophyll and nitrogen distribution in apple leaves were established.Based on the regression equation of chlorophyll and nitrogen,the chlorophyll and nitrogen of each pixel on the leaves were calculated,and the inversion map of the distribution of chlorophyll and nitrogen in apple leaves was obtained,which provided more intuitive information for the content of chlorophyll and nitrogen in apple leaves.The results showed that the inversion of chlorophyll and nitrogen distribution in apple leaves could be realized by hyperspectral imaging technology,which provided a scientific theoretical basis for effectively detecting plant nutrient distribution and rational fertilization.