应用FT-IR光谱指纹分析和模式识别技术溯源茶叶产地的研究  

DETERMINATION OF THE GEOGRAPHICAL ORIGIN OF TEA BY FT-IR SPECTROSCOPY ANALYSIS AND PATTERN RECOGNITION TECHNIQUE

作　　者：占茉莉[1,2] 李勇[2] 魏益民[2] 潘家荣[2] 钱和[1] 姚卫蓉[1] 

机构地区：[1]江南大学食品学院,江苏无锡214122 [2]中国农业科学院农产品加工研究所,北京100193

出　　处：《核农学报》2008年第6期829-833,850,共6页Journal of Nuclear Agricultural Sciences

基　　金：“十一五”科技支撑计划重点项目(2006BAK02A16);国家自然科学基金项目(30671484);国家科技研究发展计划(863计划)(2006AA10I268)

摘　　要：利用近红外光谱分析技术,对28份茶样品进行主成分和聚类分析。结果表明,浙江省龙井绿茶近红外原始光谱谱图差异较大,而不同产地龙井绿茶原始光谱间差异不甚明显。对原始光谱数学处理后对其进行主成份分析,发现在主成分空间内第1主成分得分绝大部分为正,继而对不同产地的样品进行主成分分析,西湖龙井有比较明显的主成分特征,区别于浙江龙井;"西湖龙井"主成份空间分布的离散度大于浙江各市县龙井的变异。对龙井绿茶样品进行聚类分析,得出相同产地的绿茶样品可聚为一类。初步表明应用近红外光谱分析技术可准确、快速、低廉地追溯茶叶的产地。Abstract：28 dried Longjing tea samples from Zhejiang Province in China were subjected to principal component analysis （PCA）, cluster analysis （CA）, discriminate analysis （DA） by FT-NIR spectroscopy. The results Showed that obvious spectral differences were observed. The difference of Longjing tea from different geographical origins was relatively unconspicuous, but recognizable. The near-infrared reflectance spectra of the tea are studied by the principal component analysis after mathematical treatment. It was indicated that principle component score of Longjing tea was positive and principal component of two different space distribution types expressed the characteristics of the tea samples, however, the characteristics can be defined as the particular features which distinguish Xihu Longjing tea from Zhejiang Longjing tea. And variation coefficient of characteristic parameters in spatial distribution of was concluded that applying FT-NIR fingerprint cost. Xihu Longing tea was larger than spectroscopy to trace geographical those of the other in Zhejiang province. It origin of tea is accurate, rapid with low cost.

关 键 词：茶叶 产地 近红外光谱 溯源 

分 类 号：S571.1[农业科学—茶叶生产加工]