研究目的
To develop a rapid, accurate, and non-destructive method for classifying Indonesian black tea grades using Fourier-transform near-infrared spectroscopy (FT-NIRS) and principal component analysis (PCA).
研究成果
The combination of FT-NIR spectroscopy and PCA can classify black tea grades rapidly, accurately, and non-destructively, with an accuracy level over 90%. This method is a potential alternative for tea classification in quality control processes.
研究不足
The study noted the necessity to add variations in training sets with different grades to improve the classification method. A very few samples were misclassified, indicating room for optimization in the method.
1：Experimental Design and Method Selection:
The study utilized FT-NIR spectroscopy and PCA for classifying black tea grades. The method was similar to that developed by Shou-He Yan with modifications.
2：Sample Selection and Data Sources:
Three types of tea (D1, FANN, PFI) from various estates in Indonesia were used, totaling 96 data sets divided into training (66) and validation (30) sets.
3：List of Experimental Equipment and Materials:
A near infrared spectrometer Buchi Nirfelx 500 solid? was used for spectral acquisitions. Tea samples were stored in aluminum packages and measured in vials.
4：Experimental Procedures and Operational Workflow:
Samples were mixed for homogeneity before spectral acquisition. Spectra were smoothened using the Savitszky-Golay method and analyzed using PCA.
5：Data Analysis Methods:
PCA was applied to the 2nd derivative spectra for classification, with results displayed in three-dimensional space by PC1, PC2, and PC3 axis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
