Monitoring Of Nitrogen On Summer Corn Canopy Based On Multi-Source Remote Sensing Data

Corn is one of the main food crops in China,and nitrogen has an important impact on corn growth.Reasonable and effective application of nitrogen fertilizer has an important role in promoting the formation of yield,and is conducive to achieving high yield and quality of corn.The traditional corn field information acquisition mostly relies on manual investigation,so that it is time-consuming and labor-intensive,and it is difficult to implement it on a large scale.In addition,the field data collected by personnel is quite different,and it is difficult to form a unified standard,and it will also cause different degrees of damage to corn in the field.With the rapid development of remote sensing technology,especially the application of unmanned aerial vehicles,the use of space remote sensing technology to carry out field agricultural monitoring has obvious advantages,which can provide real-time,efficient and non-destructive data acquisition technology for summer maize canopy nitrogen research.This study was conducted at the National Precision Agriculture Research Demonstration Base in Changping District,Beijing.The nitrogen maize canopy leaf nitrogen in 2012 and 2017was taken as the research object.The high-throughput crop phenotype platform was established by using near-earth hyperspectral,high-definition digital and multispectral camera carried by UAV to obtain the spectral and image data of crop growth and physical and chemical parameters.The phenotypic information of summer corn canopy nitrogen and other crops was analyzed,and the phenotypic parameters of field crops were quantitatively studied.The main contents of this paper were as follows:（1）Nitrogen analysis of canopy leaves of summer corn based on near-earth hyperspectral data with different spectral variables and algorithms.Accurate and dynamic monitoring of crop nitrogen status is the basis of scientific decisions regarding fertilization.In this study,we compared and analyzed three types of spectral variables:Sensitive spectral bands,the position of spectral features,and typical hyperspectral vegetation indices.First,the Savitzky-Golay technique was used to smooth the original spectrum,following which three types of spectral parameters describing crop spectral characteristics were extracted.Next,the successive projections algorithm（SPA）was adopted to screen out the sensitive variable set from each type of parameters.Finally,partial least squares（PLS）regression and random forest（RF）algorithms were used to comprehensively compare and analyze the performance of different types of spectral variables for estimating corn leaf nitrogen content（LNC）.The results show that the integrated variable set composed of the optimal ones screened by SPA from three types of variables had the best performance for LNC estimation by the validation data set,with the values of R²,root means square error（RMSE）,and normalized root mean square error（NRMSE）of 0.77,0.31,and 17.1%,and 0.55,0.43,and 23.9%from PLS and RF,respectively.It indicates that the PLS model with optimally multitype spectral variables can provide better fits and be a more effective tool for evaluating corn LNC.（2）Crop nitrogen analysis based on UAV multi-source imagery information fusion.Real-time,dynamic,and accurate monitoring of crop nitrogen status is of great significance for guiding fertilization-management decisions.The typical vegetation indices（VIs）and the corresponding coverage-adjusted spectral index(CASI=VIs/(1+FV_cover),where FV_coverover refers to vegetation coverage)of different crop growth stages were systematically compared and analyzed to evaluate their performances of estimating leaf nitrogen concentration（LNC）by using the high-definition digital images and multispectral images obtained synchronously by an unmanned aerial vehicle（UAV）at the National Precision Agriculture Research and Demonstration Base in 2017.First,taking into consideration the distinct soil noise due to high spatial resolution of high-definition digital images from UAV,the color space conversion method was used to remove the soil background from images and thereby extract coverage information.To characterize high-spectral-resolution multispectral images acquired synchronously by the UAV,a random forest classification method was used to eliminate some shadow noise in the images,and then the vegetation indices and corresponding CASIs were calculated.Second,the random frog algorithm was adopted to optimally select the VIs and CASIs in V12,RT,and R3 growth stages of corn.Finally,using PLS（partial least squares）regression the relationships between LNC and optimal VIs or CASIs were constructed for different growth stages,and thereby to monitor and evaluate the nitrogen nutrition of corn.The results show that CASIs provide the better estimates of LNC than VIs,and that,after removing soil background,the UAV images improve the accuracy of LNC estimates,especially when using CASIs in RT growth stage,which yields an R² of 0.59,a RMSE of 22.02%,and NRMSE（normalized root mean square error）of 8.37%.The analysis shows that the CASIs integrating information from multispectral and high-definition digital images synchronously acquired by UAV,which provide the complementary advantages of high-definition digital images with high spatial resolution and multispectral images with relatively high spectral resolution from the UAV,and the strong potential of using these indices for remote monitoring of crop nitrogen.The results provided herein also serve as useful references for other physical and chemical crop parameters for the UAV remote monitoring.In conclusion,the study used the ground hyperspectral and UAV system to obtain the summer corn canopy spectrum and images,and constructed the nitrogen quantitative inversion model.The combination of“points”and“faces”,from a single sensor to multiple types of sensors has been used to complete the timely,non-destructive and effective monitoring of summer corn canopy on different scales,which provided scientific and effective technical support for the effective monitoring of other physical and chemical parameters of crops.