CORESTA Meeting, Agronomy/Phytopathology, 2023, Cancun, APPOST 18

Study on hyperspectral multivariate linear prediction model of tobacco leaf nitrogen content

GUO Ting; LI Wujin; XIAO Xi; LI Hongguang; LI Wen
Chenzhou Tobacco Company, Chenzhou City, Hunan Province, China

The nitrogen content of tobacco leaves has a direct bearing on the gene expressions of nitrogen metabolism-related enzymes and on the amount of nitrogen metabolites (Alinat et al., 2015). Hyperspectral remote sensing, originating in the 1920s, offers an important tool for experimental science, and this technique can be used to identify molecular and atomic structures (Fan et al., 2022). Since different biochemical components of crops have distinct absorption bands (El-Naggar et al., 2021), it is feasible to monitor crop quality parameters based on optical remote sensing data. The objective of this study was to analyse the quantitative relationship between the nitrogen content of tobacco leaves and hyperspectral variables in three developmental stages and establish the hyperspectral prediction model for nitrogen content of tobacco leaves in order to obtain the nitrogen content of tobacco leaves accurately and effectively during the whole growth period. This study used the field canopy spectrum of the three critical periods of tobacco rosette stage, vigorous growth stage and topping stage. The correlation analysis of field canopy spectrum, first derivative spectrum, hyperspectral parameters and vegetation index with the nitrogen content of tobacco leaves was carried out one by one, and the prediction model was established by multiple linear regression using the variables with the best correlation coefficient. The results show that the first derivative spectrum, EVI II and green peak position show strong correlation, which is suitable for introducing multivariate equations as independent variables. Finally, the modeling determination coefficient (R2) is 0.66, RMSE is 0.40, and MAPE is 11 %. The validation results showed that R2 was 0.73, RMSE was 0.38, and MAPE was 8.33 %. It proved that this model could accurately predict the nitrogen content of tobacco leaves and could meet the requirements of large-scale statistical monitoring of tobacco quality indicators in the field.