赣南脐橙叶片含水量和叶绿素含量的近红外检测

The Measurement of Moisture Content and Chlorophyll Content in the Leaves of Citrus sinensis( L. ) Osbeck by Near Infrared

[目的]得到一种快速检测叶片含水量和叶绿素含量的定量估计模型。[方法]选取88片赣南脐橙叶片作为研究对象,运用近红外光谱技术检测叶片含水量和叶绿素含量2个指标。通过6种不同预处理方法（SG/MSC/1st D/2nd D/SNV/Baseline）在全波段4 000-12 000 cm^-1范围内建立了叶片含水量和叶绿素含量的PLS、PCR和LS-SVM定量检测模型。[结果]试验得出,在预测叶片含水量的模型中,PLS、PCR和LS-SVM预测效果整体相差不大,其最优的预测模型是使用PLS建模,其中RP=0.985,RMSEP=0.023。在预测叶片叶绿素含量的模型中,3种方法预测效果相差不大,最优预测模型是MSC-LS-VSM,其中RP=0.933,RMSEP=0.23。[结论]研究表明,应用近红外光谱技术对赣南脐橙叶片含水量和叶绿素含量的快速检测具有可行性。

[ Objective ] To obtain quantitative estimation model for rapid detection of leaf moisture content and chlorophyll content. [ Method] With 88 pieces of Citrus sinensis leaf as the research object, leaf water content and chlorophyll content was measured by near infrared technique. Quantitative model was developed using partial least squares （PLS） , principal component regression （PCR） and least squares support vector machine （LS-SVM）, and six different pretreatment methods were used in spectra processing： SG, MSC, 1st D, 2ndD, SNV, Baseline with a full spectral wave number of 4 000 - 12 000 cm^-1. [ Result] The experimental results in leaf water content prediction model show that the prediction results had little difference with PLS, PCR and LS-SVM model, and the PLS model of leaf water content gave the best results with the correlation coefficient of 0. 985, the root mean square error of prediction of 0. 023. In the prediction model of the leaf chlorophyll content, there was a small difference among 3 kinds of models, and the best prediction model was processing spectra with LS-VSM after MSC pretreatment, with correlation coefficient of 0. 933, the root mean square error of prediction of 0.23. [ Conclusion ] It shows that, application of near infrared spectroscopy for rapid detection of water content and chlorophyll content of Citrus sinensis leaves is feasible.