高光谱成像
绿茶
偏最小二乘回归
儿茶素
化学
数学
人工智能
模式识别(心理学)
生物系统
色谱法
食品科学
多酚
计算机科学
统计
生物
抗氧化剂
生物化学
作者
Feilan Li,Jingfei Shen,Qianfeng Yang,Yongning Wei,Yifan Zuo,Yujie Wang,Jingming Ning,Luqing Li
标识
DOI:10.1016/j.fochx.2024.101538
摘要
The quality of green tea deteriorates the longer it is stored. However, there is a lack of accurate and rapid methods for determining the storage period of tea. In this study, hyperspectral imaging (HSI) was used to determine the storage period of green teas stored at 4 °C (set 1) and 25 °C (set 2), and to quantify and visualize the main chemical components (e.g. catechins). In this study, three prediction algorithms were compared, in which partial least squares discriminant analysis outperformed the other models in qualitative discrimination, with 98% and 96% correct discrimination for two sets, respectively. Moreover, quantitative models for ester catechins, simple catechins, and total catechins were developed with Rp > 0.90 and RPD > 1.0, indicating that the models were reliable. Further, a more intuitive visualization of catechin content was achieved. In conclusion, the HSI provides a rapid, non-destructive method to determine the freshness of stored green tea.
科研通智能强力驱动
Strongly Powered by AbleSci AI