食品科学 ›› 2021, Vol. 42 ›› Issue (16): 155-161.doi: 10.7506/spkx1002-6630-20200315-237

• 成分分析 • 上一篇    下一篇

六堡茶茶汤色泽量化分析3 种方法的比较

马婉君,马士成,施江,龙志荣,董鹏,张悦,林智,吕海鹏   

  1. (1.农业农村部茶树生物学与资源利用重点实验室,中国农业科学院茶叶研究所,浙江 杭州 310008;2.中国农业科学院研究生院,北京 100081;3.梧州市茶产业发展服务中心,广西 梧州 543000)
  • 发布日期:2021-08-27
  • 基金资助:
    现代农业产业技术体系建设专项(CARS-19);中国农业科学院创新工程项目(CAAS-ASTIP-2014-TRICAAS); 广西六堡茶产业科研人才小高地项目

Comparative Study on Three Quantitative Analysis Methods for the Color Liupao Tea Infusion

MA Wanjun, MA Shicheng, SHI Jiang, LONG Zhirong, DONG Peng, ZHANG Yue, LIN Zhi, LÜ Haipeng   

  1. (1. Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture and Rural Affairs, Tea Research Institute, Chinese Academy of Agricultural Sciences, Hangzhou 310008, China; 2. Graduate School of Chinese Academy of Agricultural Sciences, Beijing 100081, China; 3. Wuzhou Tea Industry Development Service Center, Wuzhou 543000, China)
  • Published:2021-08-27

摘要: 为比较不同方法所得色泽参数的差异,探寻一种能够更加客观真实反映茶汤色泽的量化方法,选用36?个六堡茶散茶样品为研究对象,分别采用3?种不同方法提取茶汤的色泽参数,并对色泽参数进行模拟验证对比分析。结果表明,方法1基于Adobe?Color?CC和方法2基于Matlab的茶汤色泽量化分析方法所测定的不同茶样间色泽参数变化规律基本一致,而方法3直接采用台式色差仪所测定结果与方法1和方法2相比有明显差异,色泽参数相关性分析表明这些差异主要在于茶汤黄度值的变化不同。采用人工神经网络分析模型发现,上述3?种方法所得色泽参数对汤色得分的预测准确度分别为86.1%、83.7%和63.8%。以实际茶汤色泽为参考,通过色泽参数模拟验证对比分析发现,基于计算机视觉的图像采集装置结合Adobe?Color?CC的茶汤色泽参数提取方法能够更加真实地反映出六堡茶茶汤的实际色泽,表明该方法有望为茶汤色泽的量化分析提供一种相对客观、准确的分析检测方法。

关键词: 六堡茶;茶汤;色泽;量化分析;方法?

Abstract: This study compared the difference in color parameters measured by different methods for tea infusion in order to explore a quantitative method that can reflect the color of tea infusion more objectively and truly. In total 36 Liupao tea samples were selected for preparation of tea infusions, and three different methods were used to extract color parameters from the tea infusions. The color parameters were simulated and verified for comparative analysis. The quantitative results of method 1 based on Adobe Color CC and method 2 based on Matlab for color parameters were generally consistent with each other, but significantly different from those of method 3 (direct measurement with a bench-top colorimeter). Correlation analysis of color parameters revealed that the difference was mainly attributed to the yellowness value. The results of artificial neural network (ANN) modeling showed that the prediction accuracies of the three methods for color parameters were 86.1%, 83.7% and 63.8%, respectively. The color parameter extraction method of image acquisition based on computer vision combined with Adobe Color CC could more truly reflect the actual color of Liupao tea, suggesting that this method provides a relatively objective and accurate detection method for the quantitative analysis of tea infusion color.

Key words: Liupao tea; tea infusion; color; quantification; methods

中图分类号: