线椒货架期的近红外光谱定性判别

Qualitative discrimination of line pepper in shelf- life by near infrared spectroscopy

利用近红外漫反射光谱技术对线椒的货架期进行定性判别研究。实验以常温货架期1、3、5 d的线椒为研究对象,利用主成分分析法（PCA）建立近红外漫反射定性判别模型,在全光谱范围（400~2500 nm）内比较了不同的光谱预处理方法结合不同散射和标准化方法对所建模型的影响。结果表明,采用全光谱下Log（1/R）＋None光谱预处理方法建立的模型预测最好,该模型的交互验证相关系数（R_（CV））为0.9455,交互验证误差（SECV）为0.1534,其正确分类率达95.56%~100%,预测准确率达88.89%~97.78%,该模型能够准确地区分不同货架期的线椒鲜果。因此,近红外光谱技术为线椒货架期的鉴别提供了一种新方法。

Near-infrared diffuse reflectance spectroscopy was used for qualitative discrimination of line pepper in shelf-life,and the line pepper storaged at room temperature during shelf life（ 1,3,5 days）were as materials.The principal component analysis（PCA） was used to establish Near-infrared diffuse reflectance qualitative discriminate model.in the spectral region between 400-2500 nm,calibration results of line pepper were compared with different derivative treatments and different scatter and standard treatments.The results showed that original spectrum combined with none in the wavelength range of 400-2500 nm, provided better prediction performance of line pepper,with the cross-validation correlation coefficient（ Rcv ） of 0.9455 ,the cross-validation error（SECV） of 0.1534, the correct identification rate of discriminate model was 95.56% to 100%, predict accuracy was 88.89% to 97.78%. The model can accurately distinguish between different lines pepper fruit shelf life.So Near infrared spectroscopy could offer a new approach to the line pepper in shelf-life.