Spectral Pre-Processing Based on Convolutional Neural Network

Qing-liang JIAO; Ming LIU; Kun YU; Zi-long LIU; Ling-qin KONG; Mei HUI; Li-quan DONG; Yue-jin ZHAO

doi:10.3964/j.issn.1000-0593(2022)01-0292-06

Spectroscopy and Spectral Analysis, Volume. 42, Issue 1, 292(2022)

Spectral Pre-Processing Based on Convolutional Neural Network

Qing-liang JIAO^1、*, Ming LIU^{1、1; *;}, Kun YU^2、2;, Zi-long LIU^{2、2; 3;}, Ling-qin KONG^1、1;, Mei HUI^1、1;, Li-quan DONG^1、1;, and Yue-jin ZHAO^1、1;

Author Affiliations

¹1. Beijing Key Laboratory for Precision Optoelectronic Measurement Instrument and Technology, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China

²2. Henan Key Laboratory of Infrared Materials and Spectrum Measures and Applications, School of Physics, Henan Normal University, Xinxiang 453007, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(7)

Fig. 1. The structure of proposed CNN

Download full size

View in Article

Fig. 2. The result of spectral denoising
(a): RMSE; (b): GFC

Download full size

View in Article

Fig. 3. The result of baseline correction
(a): RMSE; (b): GFC

Download full size

View in Article

Fig. 4. Effect of denoising and baseline correction
(a): Noise and Ours; (b): DWT+[6], DWT+[8] and Ours; (c): SG+[6], SG+[8] and Ouss

Download full size

View in Article

Table 1. The result of denoising and baseline correction(RMSE\GFC)

View table

View in Article

Table 1. The result of denoising and baseline correction(RMSE\GFC)

SNR of noise	DWT+[6]	DWT+[8]	SG+[6]	SG+[8]	Ours
0	0.981\6.379	0.973\5.963	0.577\1.197	0.571\1.183	0.146\1.011
10	0.313\1.961	0.308\1.964	0.203\0.976	0.219\0.971	0.059\0.980
20	0.099\1.061	0.096\1.059	0.068\0.952	0.067\0.950	0.020\0.988
30	0.031\1.012	0.030\1.010	0.011\1.001	0.011\1.001	0.006\0.999
40	0.010\1.000	0.010\1.000	0.007\0.998	0.007\0.999	0.002\0.999

Table 2. The result of spectrum peak detection

View table

View in Article

Table 2. The result of spectrum peak detection

	标准		TD		[10]		Ours
	位置	强度	位置	强度	位置	强度	位置	强度
1	3	0.586 7	3	0.586 7	3	0.586 7	3	0.586 7
2	7	0.843 9	7	0.843 9	7	0.843 9	7	0.843 9
3	13	1.055 4	13.26	1.030 7	13.14	1.053 1	13	1.055 4
4	15	0.875 5	-	-	14.69	0.919 6	15.02	0.870 5
5	19	0.762 3	19	0.762 3	19	0.762 3	19	0.762 3

Table 3. RMSE of quantitative analysis
View table
View in Article
Table 3. RMSE of quantitative analysis
Denoising DWT SG
Baseline [6] [8] [6] [8] Ours
Peak TD [10] TD [10] TD [10] TD [10]
Moisture 0.158 0.148 0.156 0.147 0.153 0.132 0.153 0.136 0.119
Oil 0.783 0.630 0.781 0.633 0.759 0.616 0.753 0.611 0.414

Tools

Get Citation

Copy Citation Text

Qing-liang JIAO, Ming LIU, Kun YU, Zi-long LIU, Ling-qin KONG, Mei HUI, Li-quan DONG, Yue-jin ZHAO. Spectral Pre-Processing Based on Convolutional Neural Network[J]. Spectroscopy and Spectral Analysis, 2022, 42(1): 292

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites