Research Article

Qualitative Discrimination of Intact Tobacco Leaves Based on Near-Infrared Technology

Figure 3

Two-dimensional projection distribution diagrams of the training set (a, d), sample scatter diagrams of the correct discrimination of the validation set (b, e), and sample scatter diagrams of the wrong discrimination of the validation set (c, f) for the independent spectra and average spectra modeling (position-GV discrimination).
(a)
(b)
(c)
(d)
(e)
(f)