Nondestructive Identification of Different Chrysanthemum Varieties Based on Near-Infrared Spectroscopy and Pattern Recognition Methods

Jia-yi LI; Mei YU; Mai-quan LI; Yu ZHENG; Pao LI

doi:10.3964/j.issn.1000-0593(2022)04-1129-05

Journals >Spectroscopy and Spectral Analysis >Volume 42 >Issue 4 >Page 1129 > Article

Spectroscopy and Spectral Analysis
Vol. 42, Issue 4, 1129 (2022)

Nondestructive Identification of Different Chrysanthemum Varieties Based on Near-Infrared Spectroscopy and Pattern Recognition Methods

Jia-yi LI^1、*, Mei YU^1、1;, Mai-quan LI^1、1;, Yu ZHENG^{2、2; *;}, and Pao LI^{1、1; 3; *;}

Author Affiliations

¹1. College of Food Science and Technology, Hunan Agricultural University, Changsha 410128, China

²2. School of Medicine, Hunan Normal University, Changsha 410013, China

show less

DOI: 10.3964/j.issn.1000-0593(2022)04-1129-05 Cite this Article

Jia-yi LI, Mei YU, Mai-quan LI, Yu ZHENG, Pao LI. Nondestructive Identification of Different Chrysanthemum Varieties Based on Near-Infrared Spectroscopy and Pattern Recognition Methods[J]. Spectroscopy and Spectral Analysis, 2022, 42(4): 1129 Copy Citation Text

show less

Fig. 1. Original spectra of the complete (a) and powdered samples (b)

Download full size | View in the Article

Fig. 2. PCA results of the original spectra of complete (a) and powdered sample (b)

Download full size | View in the Article

Fig. 3. CWT-FLD result of the complete samples (a), FLD result of original spectra of the powdered samples (b)

Download full size | View in the Article

Dataset	Pretreatment	SIMCA/%					FLD/%
Dataset	Pretreatment	Chu	Gong	Hang	Bo	Whole	Chu	Gong	Hang	Bo	Whole
Complete	Original	100	50	60	70	70	60	100	100	90	87.5
	De-bias	100	0	70	50	55	70	100	90	80	85
	DT	90	10	60	60	55	70	100	90	90	87.5
	SNV	90	10	60	70	57.5	80	100	80	90	87.5
	Min-Max	80	60	60	100	75	60	100	90	90	85
	MSC	90	10	60	40	50	70	100	90	90	87.5
	1st	100	80	70	60	77.5	80	100	100	100	95
	2nd	100	0	60	40	50	60	100	100	90	87.5
	CWT	100	70	70	50	72.5	90	100	100	100	97.5
	1st-DT	100	80	70	50	75	70	100	100	100	92.5
	1st-SNV	100	90	100	60	87.5	90	100	100	70	90
	1st-MSC	100	90	100	90	95	80	90	100	90	90
	CWT-MSC	100	90	90	90	92.5	70	90	100	90	87.5
	CWT-SNV	90	90	100	90	92.5	80	90	100	80	87.5
Powdered	Original	70	100	100	100	92.5	100	100	100	100	100
	De-bias	70	80	100	100	87.5	100	100	100	100	100
	DT	70	70	100	80	80	100	100	100	100	100
	SNV	70	90	20	70	62.5	100	100	100	90	97.5
	Min-Max	60	80	80	90	77.5	100	100	100	90	97.5
	MSC	70	90	20	70	62.5	100	100	100	90	97.5
	1st	60	100	100	100	90	100	100	100	100	100
	2nd	70	30	60	50	52.5	80	80	70	50	70
	CWT	60	90	90	90	82.5	90	100	100	100	97.5
	1st-DT	50	90	90	100	82.5	100	100	100	90	97.5
	1st-SNV	70	50	20	70	52.5	70	80	80	80	77.5
	1st-MSC	70	60	20	80	57.5	70	80	80	70	75
	CWT-MSC	80	60	20	60	55	70	80	80	90	80
	CWT-SNV	70	60	10	60	50	60	80	80	80	75

Table 1. Identification accuracies obtained by SIMCA/FLD and pretreatment methods

Download Citation

Save the article for my favorites

Paper Information