Identification of a Grass Species Using a Terahertz Wave Based on Hybrid Machine Learning Method

Wang Fang; Zhang Chunhong; Zhao Jingfeng; Ha Sibateer; Zhang Yu

doi:10.3788/LOP202158.0330001

Journals >Laser & Optoelectronics Progress >Volume 58 >Issue 3 >Page 3300011 > Article

Laser & Optoelectronics Progress
Vol. 58, Issue 3, 3300011 (2021)

Identification of a Grass Species Using a Terahertz Wave Based on Hybrid Machine Learning Method

Wang Fang¹, Zhang Chunhong¹, Zhao Jingfeng^2、*, Ha Sibateer², and Zhang Yu¹

Author Affiliations

¹College of Science, China University of Petroleum, Beijing 102249, China

²Inner Mongolia Grassland Station, Huhhot, Inner Mongolia 010020, China

show less

DOI: 10.3788/LOP202158.0330001 Cite this Article Set citation alerts

Wang Fang, Zhang Chunhong, Zhao Jingfeng, Ha Sibateer, Zhang Yu. Identification of a Grass Species Using a Terahertz Wave Based on Hybrid Machine Learning Method[J]. Laser & Optoelectronics Progress, 2021, 58(3): 3300011 Copy Citation Text

show less

Fig. 1. Schematic of THz-TDS system

Download full size

Terahertz time and frequency domain spectral waveforms of five Astragalus adsurgens Pall. seeds. (a) Terahertz time domain spectral waveforms; (b) terahertz frequency domain spectral waveforms

Fig. 2. Terahertz time and frequency domain spectral waveforms of five Astragalus adsurgens Pall. seeds. (a) Terahertz time domain spectral waveforms; (b) terahertz frequency domain spectral waveforms

Download full size

Fig. 3. Absorption coefficient spectra of five Astragalus adsurgens Pall. seeds

Download full size

Fig. 4. Average absorption coefficient and standard deviation of five Astragalus adsurgens Pall. seeds

Download full size

$Refractive index spectra of five Astragalus adsurgens Pall. seeds$

Fig. 5. Refractive index spectra of five Astragalus adsurgens Pall. seeds

Download full size

Number	Name	Place of origin	Year
Sample 1	Sha Da Wang 1	Helin	2010
Sample 2	Sha Da Wang 2	Helin	2012
Sample 3	Sha Da Wang 3	Helin	Before 2010
Sample 4	Sha Da Wang 4	Helin	2016
Sample 5	Sha Da Wang 5	Helin	2016

Table 1. Relevant information of five samples of<i> Astragalus adsurgens</i> Pall. seeds

No.	Classification accuracy of all kinds of samples/%					Classification accuracy of five samples/%
No.	Sample 1	Sample 2	Sample 3	Sample 4	Sample 5	Classification accuracy of five samples/%
Average classification accuracy/%	81.60	85.00	86.10	81.90	90.90	85.00
1	75	100	86	67	86	83.30
2	80	89	86	80	100	86.70
3	91	60	100	86	67	83.30
4	60	100	100	67	86	83.30
5	80	67	67	100	100	83.30
6	75	100	89	80	100	86.70
7	100	63	100	100	80	86.70
8	80	100	78	71	100	83.30
9	75	85	75	88	100	86.70
10	100	83	80	80	90	86.70

Table 2. Classification results of RF model

Component	Eigenvector
Component	Eigenvalue	Variance contribution rate/%	Cumulative variance contribution rate/%
1	84.29	95.79	95.79
2	3.214	3.650	99.44
3	0.2125	0.2400	99.68

Table 3. Eigenvector of principle component

No.	Classification accuracy of all kinds of samples/%					Classification accuracy of five samples/%
No.	Sample 1	Sample 2	Sample 3	Sample 4	Sample 5	Classification accuracy of five samples/%
Average classification accuracy/%	94.20	92.10	88.30	97.00	91.40	91.20
1	80	100	100	90	100	93.30
2	100	83	89	80	100	90.00
3	100	83	75	100	100	93.30
4	100	80	83	100	75	90.00
5	100	100	80	100	71	90.00
6	86	100	83	100	100	93.30
7	86	75	100	100	100	90.00
8	90	100	100	100	100	96.70
9	100	100	87	100	88	90.00
10	100	100	86	100	80	93.30

Table 4. Classification results of PCA-RF model

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information