Figure 1.
Hyperspectral imaging.
Figure 2.
Schematic diagram of the NIR hyperspectral imaging system.
Table 1.
Statistical information of calibration and prediction sets.
Figure 3.
Spectral reflectance curves of the four different varieties of sesame oil.
Figure 4.
The changing trend of the number of sampled variables (a), 10-fold RMSECV values (b) and regression coefficients of each variable (c) with the increasing of sampling runs.
The line (marked by asterisk).
Table 2.
Effective wavelengths suggested by CARS.
Figure 5.
Effective wavelengths selected by SPA.
Figure 6.
Effective wavelengths selected by x-LW.
Figure 7.
(a) Correct classification rate of each model at each selected wavelength suggested by SPA, (b) Correct classification rate of each model at each selected wavelength suggested by x-LW.
Table 3.
Correct classification rate of different models based on different wavelengths selection methods.