Spatial distribution of total polyphenols in multi-type of tea using near-infrared hyperspectral imaging

Polyphenols are the key taste and health components of tea. Typically, chemical analysis is used to determine the total polyphenol (TP) content of tea. However, this process is time consuming. In this study, the TP content of six tea types, namely green, white, yellow, oolong, black, and dark, was assessed using near-infrared hyperspectral imaging. Here, 100% accuracy was achieved for both the calibration and prediction sets for qualitative discrimination of tea by using the principal component analysis-K-nearest neighbor model. Important wavelengths of TP were selected using the regression coefficients (RCs) of the partial least squares regression (PLSR) model. The proposed RC-PLSR model yielded satisfactory prediction results with a residual predictive deviation of 3.34. The differences in the spatial distribution of TP in various tea samples were visualized using distribution maps. This study provided a rapid and nondestructive method for the identification of tea types and the visualization of the TP content of tea.

Automated summary

Polyphenols are key components in tea for taste and health benefits. This study used near-infrared hyperspectral imaging to assess TP content in six types of tea, achieving high accuracy in discriminating tea types using a PCA-KNN model. Selected wavelengths through PLSR model helped predict TP content effectively, and distribution maps visualized spatial differences in TP among different tea samples. This study provided a rapid and nondestructive method for tea type identification and TP content visualization.