Discrimination of black tea fermentation degree based on multi-data fusion of near-infrared spectroscopy and machine vision

Fermentation is a key step in the production of black tea and has an important impact on the quality of black tea. In this research, the fermentation degree of black tea was evaluated by a portable near-infrared spectrometer and a charge-coupled device camera. A total of 180 samples of black tea were taken at a variety of periods during the fermentation process, and their near-infrared spectra and images were measured. After the analyses of the changes in tea polyphenol and catechin contents measured by ultraviolet spectrophotometry and high-performance liquid chromatography, the fermentation degree for black tea was divided into three stages. Discrimination models based on spectra, images and their data fusion were established through linear discriminant analysis (LDA), random forest (RF) and support vector machine (SVM). Among them, the discrimination model established by successive projections algorithm (SPA) extraction of spectral variables and Pearson correlation analysis extraction of image variables obtained satisfactory results with 100.00% and 95.00% accuracies of the calibration set and prediction set, respectively. The study demonstrated that the middle-level data fusion of near-infrared spectroscopy and machine vision could be employed as a rapid and nondestructive technique to discriminate the fermentation degree of black tea.