Tea Analyzer: A low-cost and portable tool for quality quantification of postharvest fresh tea leaves

The quality components in postharvest tea leaves determine their quality characteristics and processing suitability. Currently, chemical analysis is widely used to determine these indicators. However, this method is timeconsuming and laborious. In this study, we developed a NIR-based analytical tool on a smartphone for in situ and rapid detection of quality indicators in postharvest tea leaves of multiple varieties and harvesting standards. The smartphone and the miniature NIR spectrometer are used for data acquisition and transfer to the phone via Bluetooth. Partial least squares (PLS) modeling coupled with spectral pre-processing and selection of characteristic wavelengths were used to obtain the optimal predictive model. The results showed that the quality of various tea tree varieties and harvesting standards differed considerably. The PLS models with suitable wavelengths achieved acceptable predictive performance for tea polyphenols, amino acids, and polyphenol to amino acid ratio (P/A values). The determination coefficients of the prediction set for tea polyphenols, amino acids, and the P/A value were 0.90, 0.91, and 0.91, and residual predictive deviations were 2.24, 2.43, and 2.42, respectively. Based on the optimal model constructed, a smartphone based analytical software was developed to achieve low-cost and rapid quality quantification of postharvest fresh tea leaves.

Automated summary

In this study, a NIR-based smartphone analytical tool was developed for rapid detection of quality indicators in postharvest tea leaves, showing significant quality differences among different tea tree varieties and harvesting standards. PLS models based on suitable wavelengths achieved acceptable predictive performance for key indicators.