Hitherto, the intelligent evaluation of black tea fermentation is still an unsolved problem because it is difficult to obtain the complicated changes information of tea composition, color, texture and aroma in the fermentation process at the same time. In this research, hyperspectral imaging technology was used to collect sensory information including taste (sample spectra), vision (sample color image) and olfactory (pH, porphyrin and metalloporphyrin (TPP) sensing array spectra) of fermentation leaves. Subsequently, different data fusion strategies combined with support vector machine algorithm (SVM) were used to establish the fermentation degree discrimination model. The performance of the established models using data fusion strategy were better than that of the model using each single information. The middle-level-PCA strategy achieved a satisfactory performance, with the variable compression rate of 99% and the accuracy of 95% for the prediction set. Remarkably, for the most important moderate fermentation class, the precision and recall of the model were 100% both in calibration and prediction set. These results demonstrated that our proposed strategy could accurately evaluate the fermentation degree of black tea. Display omitted •A self-built olfactory sensor including 1 pH and 3 TPP were developed.•Hyperspectral imaging technology was used to obtain sensory information of samples.•Different data fusion strategies were used to evaluate fermentation degree of black tea.