The production of high-quality tea by Camellia sinensis (L.) O. Ktze is the goal pursued by both producers and consumers. Rapid, nondestructive, and low-cost monitoring methods for monitoring tea quality could improve the tea quality and the economic benefits associated with tea. This research explored the possibility of monitoring tea leaf quality from multi-spectral images. Threshold segmentation and manual sampling methods were used to eliminate the image background, after which the spectral features were constructed. Based on this, the texture features of the multi-spectral images of the tea canopy were extracted. Three machine learning methods, partial least squares regression, support vector machine regression, and random forest regression (RFR), were used to construct and train multiple monitoring models. Further, the four key quality parameters of tea polyphenols, total sugars, free amino acids, and caffeine content were estimated using these models. Finally, the effects of automatic and manual image background removal methods, different regression methods, and texture features on the model accuracies were compared. The results showed that the spectral characteristics of the canopy of fresh tea leaves were significantly correlated with the tea quality parameters (r ≥ 0.462). Among the sampling methods, the EXG_Ostu sampling method was best for prediction, whereas, among the models, RFR was the best fitted modeling algorithm for three of four quality parameters. The R² and root-mean-square error values of the built model were 0.85 and 0.16, respectively. In addition, the texture features extracted from the canopy image improved the prediction accuracy of most models. This research confirms the modeling application of a combination of multi-spectral images and chemometrics, as a low-cost, fast, reliable, and nondestructive quality control method, which can effectively monitor the quality of fresh tea leaves. This provides a scientific reference for the research and development of portable tea quality monitoring equipment that has general applicability in the future.

Chlorophyll and nitrogen contents were used as leaf physiological parameters. Based on multispectral images from multiple detection angles and the stoichiometric data of tea (Camellia sinensis) leaves in different positions on plants, the spatial differences in tea physiological parameters were explored, and the full channel difference vegetation index was established to effectively remove soil and shadow noise. Support vector machine, random forest (RF), partial least square, and back-propagation algorithms from the multispectral images of leaf and canopy scales were then used to train the tea physiological parameter detection model. Finally, the detection effects of the multispectral images obtained from different angles on the physiological parameters of the top, middle, and bottom tea leaves were analysed and compared. The results revealed distinct spatial differences in the physiological parameters of tea leaves in individual plants. Chlorophyll content was lowest at the top and relatively high at the middle and bottom; nitrogen content was the highest at the top and relatively low at the middle and bottom. The horizontal distribution of physiological parameters was similar, i.e., the values in the east and south were high, whereas those in the west and north were low. The multispectral detection accuracy of the physiological parameters at the leaf scale was better than that at the canopy scale; the model trained by the RF algorithm had the highest comprehensive accuracy. The coefficient of determination between the predicted and measured values of the spad-502 plus instrument was (R²) = 0.79, and the root mean square error (RMSE) was 0.11. The predicted result for the nitrogen content and the measured value was R² = 0.36 and RMSE = 0.03. The detection accuracy of the multispectral image taken at 60° for the physiological parameters of tea was generally superior to those taken at other shooting angles. These results can guide the high-precision remote sensing detection of tea physiological parameters.