Raman and FT-IR Spectroscopy Coupled with Machine Learning for the Discrimination of Different Vegetable Crop Seed Varieties

Abstract

The aim of this research is to investigate the potential of Raman and FT-IR spectroscopy as well as mathematical linear and non-linear models as a tool for the discrimination of different seed varieties of paprika, tomato, and lettuce species. After visual inspection of spectra, pre-processing was applied in the following combinations: (1) smoothing + linear baseline correction + unit vector normalization; (2) smoothing + linear baseline correction + unit vector normalization + full multiplicative scatter correction; (3) smoothing + baseline correction + unit vector normalization + second-order derivative. Pre-processing was followed by Principal Component Analysis (PCA), and several classification methods were applied after that: the Support Vector Machines (SVM) algorithm, Partial Least Square Discriminant Analysis (PLS-DA), and Principal Component Analysis-Quadratic Discriminant Analysis (PCA-QDA). SVM showed the best classification power in both Raman (100.00, 99.37, and 92.71% for lettuce, paprika, and tomato varieties, respectively) and FT-IR spectroscopy (99.37, 92.50, and 97.50% for lettuce, paprika, and tomato varieties, respectively). Moreover, our novel approach of merging Raman and FT-IR spectra significantly contributed to the accuracy of some models, giving results of 100.00, 100.00, and 95.00% for lettuce, tomato, and paprika varieties, respectively. Our results indicate that Raman and FT-IR spectroscopy coupled with machine learning could be a promising tool for the rapid and rational evaluation and management of genetic resources in ex situ and in situ seed collections.

Keywords: breeding; merging spectra; support vector machine (SVM); vegetable seed; vibrational spectroscopy.

Figure 1

Baseline corrected and smoothed Raman spectra of paprika (A), tomato (B), and lettuce (C); AT—Atrakcija, MK—Majska kraljica, GL—Great Lake, LL—Ljubljanska ledenka, PCH—Palanačko čudo, STR—Strižanka, PK—Palanačka kapija, DF—Delfina, CS—Crvena Stena, L—Lider F1, K—King F1, VS—Volovsko srce.

Figure 2

Baseline corrected and smoothed FT-IR spectra of paprika (A), tomato (B), and lettuce (C); AT—Atrakcija, MK—Majska kraljica, GL—Great Lake, LL—Ljubljanska ledenka, PCH—Palanačko čudo, STR—Strižanka, PK—Palanačka kapija, DF—Delfina, CS—Crvena Stena, L—Lider F1, K—King F1, VS—Volovsko srce.

Figure 3

PCA score plots (PC1 vs. PC2) of lettuce varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD. AT—Atrakcija, MK—Majska kraljica, GL—Great Lake, LL—Ljubljanska ledenka.

Figure 4

PCA loading plots of lettuce varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD.

Figure 5

PCA score plots (PC1 vs. PC2) of paprika seed varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD. PCH—Palanačko čudo, STR—Strižanka, PK—Palanačka kapija, DF—Delfina.

Figure 6

PCA loading plots of paprika seed varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD.

Figure 7

PCA score plots (PC1 vs. PC2) of tomato varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD. CS—Crvena Stena, L—Lider F1, K—King F1, VS—Volovsko srce.

Figure 8

PCA loading plots of tomato seed varieties: Raman spectra combined with (A) BC+N, (B) BC+N+MSC, and (C) BC+N+MSC+2nd OD; FT-IR spectra with (D) BC+N, (E) BC+N+MSC, (F) BC+N+MSC+2nd OD.

Figure 9

Preprocessed and merged Raman and FT-IR spectra; (A) lettuce: ATR—Atrakcija, MK—Majska kraljica, GL—Great Lake, LL—Ljubljanska ledenka; (B) tomato: CS—Crvena Stena, L—Lider F1, K—King F1, VS—Volovsko srce; (C) paprika: PCH—Palanačko čudo, STR—Strižanka, PK—Palanačka kapija, DF—Delfina.

Figure 10

PCA score plots (PC1 vs. PC2) of lettuce varieties: Raman+FT-IR spectra combined with BC+N (A), BC+N+MSC (B), and BC+N+MSC+2nd OD (C); tomato varieties: Raman+FT-IR spectra with BC+N (D), BC+N+MSC (E), and BC+N+MSC+2nd OD (F); paprika varieties: Raman+FT-IR spectra with BC+N (G), BC+N+MSC (H), and BC+N+MSC+2nd OD (I). ATR—Atrakcija, MK—Majska kraljica, GL—Great Lake, LL—Ljubljanska ledenka; CS—Crvena Stena, L—Lider F1, K—King F1, VS—Volovsko srce; PCH—Palanačko čudo, STR—Strižanka, PK—Palanačka kapija, DF—Delfina.

Figure 11

Results of average classification accuracy performed on lettuce seed varieties: PLS-DA (A), PCA-QDA (B), SVM (C); paprika seed varieties: PLS-DA (D), PCA-QDA (E), SVM (F); tomato seed varieties: PLS-DA (G), PCA-QDA (H), SVM (I). Different small letters at each figure (A–I) implies statistical significant differences (p < 0.05).

Figure 12

Seed material of different varieties (first row-paprika; second row-lettuce; third row-tomato).