首页 | 本学科首页   官方微博 | 高级检索  
     检索      

茶汤光谱指纹图谱判别分析及相似度评价在绿茶分级中的应用
引用本文:王淑慧,曹学丽.茶汤光谱指纹图谱判别分析及相似度评价在绿茶分级中的应用[J].北京工商大学学报(自然科学版),2019,37(4):84-88.
作者姓名:王淑慧  曹学丽
作者单位:北京工商大学 北京食品营养与人类健康高精尖创新中心, 北京 100048,北京工商大学 北京食品营养与人类健康高精尖创新中心, 北京 100048
摘    要:采用色差仪对茶汤汤色进行400~700nm连续波长扫描的方法,分别建立了7个等级信阳毛尖茶汤色光谱指纹图谱。基于各波长下的特征反射率与透射率比值,采用判别分析技术对不同等级信阳毛尖茶进行了很好的等级区别,判别正确率均大于90%;通过7种方法对不同等级茶汤光谱指纹图谱的相似度进行评价,发现各相似度计算方法均能不同程度反映茶叶等级间的差异大小,其中以新改良程度相似度法最佳。相似度评价方法与判别分析结果一致,且可实现对等级差异的量化分析,为绿茶品质评价和质量监控提供了一种客观可量化的方法。

关 键 词:信阳毛尖茶    汤色    光谱指纹图谱    判别分析    相似度评价
收稿时间:2018/12/29 0:00:00

Application of Discriminant Analysis and Similarity Evaluation of Visible Spectrum Fingerprint in Green Tea Grading
Institution:Beijing Advanced Innovation Center for Food Nutrition and Human Health, Beijing Technology and Business University, Beijing 100048, China and Beijing Advanced Innovation Center for Food Nutrition and Human Health, Beijing Technology and Business University, Beijing 100048, China
Abstract:The visible spectrum fingerprints of infusion color of seven grades of Xinyang Maojian tea were established by using the colorimeter to scan the tea infusion continuously under the wavelength of 400-700nm. Based on the ratio of reflectance and transmittance under each wavelength, the method of discriminant analysis was used to differentiate grades of Xinyang Maojiao tea with the accuracy more than 90%. Seven methods were used for fingerprint similarity calculation which all can reflect the difference between grades of the tea to different extent, while the new improved extent similarity method was proved to be the best. The results of similarity calculation in grading of teas were consistent with that of discriminant analysis and the difference between grades can be quantified. This provides another objective and quantitative method for quality evaluation and quality control of green tea.
Keywords:Xinyang Maojian tea  color of tea infusion  spectrum fingerprint  discriminant analysis  similarity evaluation
点击此处可从《北京工商大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《北京工商大学学报(自然科学版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号