Study On Nutrient Detection Of Tobacco Growing Soil Based On Near Infrared Spectroscopy

Tobacco-growing soil contains many nutrients needed for the normal growth and development of tobacco leaves,and the detection of these nutrients is an important content to ensure the yield and quality of tobacco leaves.Compared with the traditional chemical analysis methods for detecting tobacco-growing soil nutrients,near infrared spectroscopy has the advantages of faster speed,less pollution and lower cost.In this paper,tobacco-growing soils in Kunming,Chuxiong and Honghe regions of Yunnan Province were collected as the research objects,and the total nitrogen,total potassium,organic matter and total phosphorus in tobacco-growing soils were predicted by near-infrared spectroscopy combined with chemometrics.The main research contents and The result is as follows:（1）Various preprocessing methods such as standard normal variable transfor-mation（SNV）,multivariate scattering correction（MSC）,first derivative,second derivative,SG,SNV+first derivative and other pretreatment methods were used to process the spectrum of tobacco-growing soil.The difference analysis of the spectral images shows that except for SG smoothing,the spectral morphology after preprocessing will change significantly.（2）Discriminant analysis method was used to establish the qualitative analysis model of tobacco-growing soil origin discrimination.The results showed that when SNV+first derivative+SG smoothing method was used for pretreatment,the main component score of the model was selected 10,and the spectral band was 6000～10000cm^-1,the discrimination effect of tobacco-growing soil origin based on the model was better,and the prediction accuracy was 98.89%.（3）Based on the stepwise multiple linear regression（SMLR）,principal component analysis（PCA）and partial least squares（PLS）,taking the correlation coefficient（R）,the corrected root mean square error（RMSEC）and the root mean square error of prediction（RMSEP）as the model evaluation indexes,the prediction model of nutrient content in tobacco-growing soil was established by using near infrared spectroscopy analysis technology.Through the comparative analysis of the R,RMSEC and RMSEP values of each model,the near-infrared spectroscopy of tobacco-growing soil established by PLS method and the prediction effect of total nitrogen,total potassium,organic matter and total phosphorus content models were ideal.The correlation coefficient R values of each model reached 0.933,0.937,0.891 and 0.835,respectively,indicating that the established model had good prediction accuracy for the main chemical composition content of tobacco-growing soil.The desktop near infrared spectroscopy technology could facilitate the rapid analysis of soil nutrients.（4）The prediction model of near infrared spectrum and soil nutrient content of tobacco-growing soils was established by hand-held near-infrared spectroscopy.The correlation coefficients R of the prediction model of total nitrogen,total potassium,organic matter and total phosphorus content reached 0.674,0.837,0.661 and 0.603,respectively.The reliability of the prediction results was far lower than that of the desktop near-infrared spectroscopy,indicating that there were certain limitations in the rapid detection of soil nutrient content of tobacco planting by hand-held near-infrared spectroscopy,and the detection accuracy of the equipment and the reliability of the model need to be improved.To sum up,the use of near-infrared spectroscopy to detect nutrients in tobacco-growing soil,the spectral images after different pretreatment methods have different shapes,which have a great impact on the performance of the model.Using desktop near-infrared spectroscopy can detect total nitrogen,total potassium,organic matter and total phosphorus.The results provide a reference for the application of near-infrared spectroscopy to the detection of nutrient content in tobacco-growing soil.