基于图像特征提取的英红九号发酵程度判断研究OA

The fermentation degree of black tea directly affects its quality.Traditional assessment methods rely heavily on human experience,resulting in strong subjectivity and difficulty in standardization.To achieve objective and intelligent monitoring of the fermentation process for Yinghong No.9 black tea,this study proposed a method based on image feature extraction to quantify fermentation progress.Specifically,the fermentation degree was modeled as a continuous regression problem,introducing"fermentation degree"as a quantitative indicator that enabled the output of a continuous curve reflecting the dynamic changes during fermentation.High-resolution macro images of tea leaves were continuously captured throughout the fermentation process using a macro camera.Multi-dimensional color features were extracted from RGB,HSV,and Lab color spaces.Subsequently,multiple machine learning models—including Random Forest(RF),Gradient Boosting Regression Trees(GBRT),Support Vector Machine(SVM),and Artificial Neural Network(ANN)—were constructed to predict the fermentation degree of Yinghong No.9 black tea.Experimental results show that the ANN model achieved the best performance in testing and validation,with a coefficient of determination(R2)of 0.71 and a root mean square error(RMSE)of 0.17,outperforming other models and demonstrating strong generalization capability and application potential.This study provides an effective technical pathway for intelligent monitoring of black tea fermentation processes.