A Nondestructive Identification Method of Producing Regions of Citrus Based on Near Infrared Spectroscopy

Xin-xin ZHANG; Shang-ke LI; Pao LI; Yang SHAN; Li-wen JIANG; Xia LIU

doi:10.3964/j.issn.1000-0593(2021)12-3695-06

Journals >Spectroscopy and Spectral Analysis >Volume 41 >Issue 12 >Page 3695 > Article

Spectroscopy and Spectral Analysis
Vol. 41, Issue 12, 3695 (2021)

A Nondestructive Identification Method of Producing Regions of Citrus Based on Near Infrared Spectroscopy

Xin-xin ZHANG^1、*, Shang-ke LI^1、1;, Pao LI^{1、1; 2; *;}, Yang SHAN^2、2;, Li-wen JIANG^1、1;, and Xia LIU^1、1;

Author Affiliations

¹1. College of Food Science and Technology, Hunan Provincial Key Laboratory of Food Science and Biotechnology, Hunan Agricultural University, Changsha 410128, China

²2. Hunan Agricultural Product Processing Institute, Hunan Academy of Agricultural Sciences, Changsha 410125, China

show less

DOI: 10.3964/j.issn.1000-0593(2021)12-3695-06 Cite this Article

Xin-xin ZHANG, Shang-ke LI, Pao LI, Yang SHAN, Li-wen JIANG, Xia LIU. A Nondestructive Identification Method of Producing Regions of Citrus Based on Near Infrared Spectroscopy[J]. Spectroscopy and Spectral Analysis, 2021, 41(12): 3695 Copy Citation Text

show less

Fig. 1. The raw spectra
(a)—(h): equator 1—4, top, bottom, average of 4 points, average of 6 points, respectively

Download full size | View in the Article

PCA results of 4 equator points average spectra with single pretreatment methods(a)—(i): raw spectra, DT, De-bias, MSC, Min-Max, SNV, 1st, 2nd and CWT, respectively

Fig. 2. PCA results of 4 equator points average spectra with single pretreatment methods
(a)—(i): raw spectra, DT, De-bias, MSC, Min-Max, SNV, 1st, 2nd and CWT, respectively

Download full size | View in the Article

Fig. 3. Cumulative variance contribution rates at different positions
(a): 4 points on equator; (b): 4 points on equator+stem+top

Download full size | View in the Article

Fig. 4. PCA-FLD results with De-bias method
(a): 4 points average spectra; (b): 6 points average spectra

Download full size | View in the Article

Spectral acquisition site	Pretreatment	Identification accuracies/%
Spectral acquisition site	Pretreatment	Yunnan	Hunan	Wuming	Laibin	All
4 points on equator	Raw spectra	90	100	100	100	97.5
	DT	90	100	100	100	97.5
	De-bias	100	100	100	100	100
	SNV	90	100	100	90	95
	Min-Max	90	100	100	90	95
	MSC	100	100	100	100	100
	1st	80	100	100	100	95
	2nd	70	80	90	30	67.5
	CWT	80	100	100	100	95
	1st-DT	80	100	100	100	95
	1st-SNV	80	100	100	100	95
	1st-MSC	90	100	100	100	97.5
	CWT-SNV	90	100	100	100	97.5
	CWT-MSC	80	100	100	100	95
4 points on equator and the top and bottom of orange	Raw spectra	100	100	100	100	100
	DT	100	100	100	100	100
	De Bias	100	100	100	100	100
	SNV	100	100	100	100	100
	Min-Max	100	100	100	100	100
	MSC	100	100	100	100	100
	1st	100	100	110	100	100
	2nd	100	80	80	80	100
	CWT	100	100	100	90	85
	1st-DT	80	100	100	90	97.5
	1st-SNV	80	100	100	90	92.5
	1st-MSC	100	100	100	100	92.5
	CWT-SNV	80	100	100	90	100
	CWT-MSC	80	100	100	90	92.5

Table 1. Identification accuracies by PCA-FLD with different pretreatment methods

Download Citation

Save the article for my favorites

Paper Information