Fig 1.
Flowchart of main data-processing procedures to predict FAA with hyperspectral images.
Fig 2.
The hyperspectral imaging system.
1 dark room; 2 CCD camera; 3 imaging spectrograph; 4 lens; 5 light source; 6 Sample stage; 7 mobile platform; 8 mobile platform controller; 9 fiber; 10 light source controller; 11 computer.
Fig 3.
Full spectrum of yellow tea preprocessed by SG smoothing filter.
Fig 4.
The selected feature wavelengths after SPA algorithm.
Table 1.
Performance of GA-SVR models based on different data for prediction of FAA.
Fig 5.
GA-SVR model between measured and predicted free amino acids based on fusion of different data.