Research Progress of Hyperspectral Imaging in Nondestructive Testing of Vegetable Traits

Jiekai Yang; Zhiqiang Guo; Yuan Huang

doi:10.3788/LOP202259.1000002

Journals >Laser & Optoelectronics Progress >Volume 59 >Issue 10 >Page 1000002 > Article

Laser & Optoelectronics Progress
Vol. 59, Issue 10, 1000002 (2022)

Research Progress of Hyperspectral Imaging in Nondestructive Testing of Vegetable Traits

Jiekai Yang¹, Zhiqiang Guo¹, and Yuan Huang^2、*

Author Affiliations

¹School of Information Engineering, Wuhan University of Technology, Wuhan 430070, Hubei , China

²Key Laboratory of Horticultural Plant Biology, Ministry of Education, College of Horticulture & Forestry Sciences, Huazhong Agricultural University, Wuhan 430070, Hubei , China

show less

DOI: 10.3788/LOP202259.1000002 Cite this Article Set citation alerts

Jiekai Yang, Zhiqiang Guo, Yuan Huang. Research Progress of Hyperspectral Imaging in Nondestructive Testing of Vegetable Traits[J]. Laser & Optoelectronics Progress, 2022, 59(10): 1000002 Copy Citation Text

show less

Fig. 1. Schematic of different imaging methods. (a) NIRS; (b) MSI; (c) HSI

Download full size

Fig. 2. Schematic of HSI system

Download full size

Fig. 3. Schematic of band feature selection

Download full size

Fig. 4. Schematic of band feature extraction

Download full size

Fig. 5. Schematic of HSI image block construction

Download full size

Internal quality	Vegetable	Wavelength range /nm	Number of characteristic wavelengths	method Modeling	Accuracy	Reference
Hardness	Cucumber	371.05-1023.82	25	BP	91.67%	［14］
Water	Cucumber	371.05-1023.82	20	BP	95.00%	［14］
Hardness	Tomato	865.11-1711.71	47	PLSR	$R_{P}$ =0.9685； RMSEP is 0.0040	［15］
Water	Tomato	871.60-1766.30		PLSR	$R_{p}^{2}$ =0.902	［16］
DMC	Potato	900-1700	8	PSO-SVM	$R_{p}^{2}$ =0.944； RMSEP is 0.155	［19］
DMC	Potato	382-1010	22	PLSR	$R_{p}^{2}$ =0.849； RMSEP is 0.878%	［20］
SSC	Tomato	900-1700	53	PLSR	$R_{p}^{2}$ =0.8878； RMSEP is 0.6276	［21］
SSC	Hami melon	400-1000	28	SVM	$R_{p}^{2}$ =0.9404； RMSEP is 0.4027	［22］

Table 1. Application of HSI in the detection of internal quality of vegetables

Nutrient element	Vegetable	Wavelength range /nm	Number of characteristic wavelengths	Modeling method	Accuracy	Reference
N	Beet	350-1830		PCR	$R_{p}^{2}$ =0.633； RMSEP is 2.34	［25］
N	Beet	383-1003	5	SVM	$R_{p}^{2}$ =0.78； RMSEP is 3.08	［26］
N	Tomato	390.8-1050.1	7	PLSR	$R_{p}^{2}$ =0.94； RMSEP is 0.48	［27］
P	Rape	350-2500	6	PLSR	$R_{p}^{2}$ =0.769； RMSEP is 0.048%	［29］
P	Cucumber	730-1300		BP-ANN	100%	［30］
K	Cucumber	730-1300		LDA	95.00%	［32］
N，P，K	Rape	350-2500	10	Framework	80.76%	［33］
N，P，K	Tomato	390-1050	12	ANN	$R_{p}^{2}$ =0.9651； RMSEP is 0.19（N）	［34］
N，P，K	Tomato	390-1050	12	ANN	$R_{p}^{2}$ =0.9216； RMSEP is 0.33（P）	［34］
N，P，K	Tomato	390-1050	12	ANN	$R_{p}^{2}$ =0.9353； RMSEP is 0.29（K）	［34］

Table 2. Application of HSI in the monitoring of vegetable nutrient element

Vegetable diseases and insect pests	Wavelength range /nm	Number of characteristic wavelengths	Modeling method	Accuracy	Reference
Cucumber （four diseases）	400-720	12	LDA	100%	［38］
Cucumber downy mildew	400-1000	47	LS-SVM	100%（2-12 d）	［39］
Cucumber angular leaf spot	380-1030		PLSR	$R_{p}^{2}$ =0.816； RMSEP is 11.235	［40］
Tomato early blight	874-1734	14	LS-SVM	100%	［41］
Tomato fungal disease	380-1023	5	SVM	100%	［42］
Tomato botrytis	400-1030	7	LS-SVM	100%	［43］
Potato early blight	375-1018	3	LS-SVM	100%	［44］
Potato late blight	375-1018		LS-SVM	94.87%	［45］
Potato ring rot	980-1650	9	MD-LDA	100%	［46］

Table 3. Application of HSI in the diagnosis of vegetable diseases

Jiekai Yang, Zhiqiang Guo, Yuan Huang. Research Progress of Hyperspectral Imaging in Nondestructive Testing of Vegetable Traits[J]. Laser & Optoelectronics Progress, 2022, 59(10): 1000002

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information