近红外光谱法快速鉴别真伪平谷大久保桃

Rapid True or False Identification on Pinggu Okubao Peach by Near-infrared Spectroscopy

［目的］实现对真伪平谷大桃的快速鉴别。[方法］选用65个当地果园的平谷大久保桃和115个其他产地大久保桃进行试验，使用傅立叶近红外光谱仪对样品进行吸收光谱扫描，采用OPUS化学计量学分析软件对样品的光谱进行处理。［结果］通过对光谱全波段的考察，发现在7500-4000cm-1波段范围信息量较丰富，区别较大，选取此光谱区域进行后续处理。经一阶导数＋矢量归一化预处理，利用因子化法、合格性测试和主成分分析法（PCA）建立平谷大久保桃的鉴别模型，结果表明所建立的3种模型都能准确鉴别真伪平谷大久保桃，鉴别正确率高于95%。［结论］采用近红外光谱结合模式识别技术可快速、准确地鉴别平谷大久保桃的真伪。

[ Objective ] The aim was to come true the rapid identification of Pinggu peach. [ Method ] The 65 Pinggu Okubao Peaches from local gardens and 115 Big Peaches from other producing areas were selected for experiment. The Fourier near infrared spectra was used for absorption scanning and OPUS chemometrics analysis software was used to dispose sample spectrum. [ Result ] Through observation on whole band spectrum, it found that the information in the range of 7 500-4 000 cm- 1 was more rich and their difference was larger, so the spectral region was selected for subsequent treatment. After pretreatment of first derivative and vector normalization, an identification model for Pinggu Okubao Peach based on factor analysis, qualification testing and principal component analysis （PCA） was established. The result showed that the 3 established models all could identify Pinggu Okubao Peach accurately and the identification accuracy was more than 95%. [Conclusion] The near-infrared spectroscopy pattern recognition technology could identify Pinggu Okubao Peach rapidly and accurately.