可见/近红外光估测完整柑橘水果可溶性固形物含量的研究

研究用可见/近红外光谱(Vis/NIR spectroscopy)漫透射方式对柑橘类水果的可溶性固形物含量(SSC)进行了无损、快速定量分析.通过自行设计的Vis/NIR光谱系统测定了110个柑橘样品的SSC.82个柑橘样品用来建模,其余28个用来验证模型的性能.对实验室测得的柑橘水果的SSC实际值与Vis/NIR光谱数据进行相关性分析,用光谱定量分析软件中集成的偏最小二乘回归法(PLS)和主成分回归法(PCR)建立校正和预测模型.对比了不同光谱预处理方法,如微分处理,标准归一化处理(SNV),多元散射校正(MSC)和Savitzky-Golay光谱平滑方法)对模型检测结果的影响.根据预测平方根误差(RMSEP)和相关系数(r2)进行不同模型的预测性能评价,建立的最好的柑橘水果SSC预测模型的RMSEP=0.538%,r2=0.896.结果表明Vis/NIR可以作为一种准确、快速的无损检测方法来评价柑橘类水果的可溶性固形物含量.

Estimation of Soluble Solids Content of Intact Citrus Fruit by Vis/NIR Spectroscopy

Visible/Near-infrared(Vis/NIR) spectroscopy has become a very popular technique for the non-invasive assessment of intact fruit.The feasibility of using Vis/NIR spectroscopic technology for rapid quantifying soluble solids content(SSC) of citrus fruit was investigated by means of spectral transmittance mode.A total of 110 citrus fruit samples were used to develop the calibration and prediction models.The relationship between actual SSC and Vis/NIRS spectra of citrus fruit samples was analyzed via pricipal component regression(PCR) and partial least squares(PLS) regression method using TQ 6.2 spectral analysis software.Models based on the different spectral pre-processing methods were compared in the present research.Performance of different models was assessed in terms of root mean square errors of prediction(RMSEP) and correlation coefficients(r2) of prediction set of samples.The best predictive models feature a RMSEP of 0.538% and correlation coefficient(r2) of 0.801 for SSC.The results show that the Vis/NIR transmittance technique is a feasible,accurate and fast method for non-invasive estimation of citrus fruit SSC.