Storage life prediction and quality discrimination of instant green tea: Integrating computer vision, electronic nose, and electronic tongue.

Tea storage is a critical determinant in determining the quality of tea products. This study systematically investigated the quality alterations of instant green tea during storage and developed an intelligent evaluation method by integrating computer vision, electronic nose, and electronic tongue with machine learning. Quantitative chemical profiling established statistically significant correlations between conventional quality indicators and multi-sensor intelligent features. Machine learning models effectively classified the storage duration of instant green tea, with the electronic tongue achieving a classification accuracy exceeding 98 % for storage time prediction. Furthermore, data fusion combined with feature selection algorithms enhanced the predictive accuracy for both storage duration and key quality content. The integration of intelligent sensing technologies provides a robust methodology for rapid shelf-life prediction and quality discrimination of instant tea, establishing a scientific foundation for quality control and authenticity assurance in the tea industry.