Positioning of Characteristic Spectral Peaks Based on Improved Sine Cosine Algorithm

Kun Yu; Qingliang Jiao; Zilong Liu; Yiqin Jiang; Qiaoxiang Zhang; Yufang Liu

doi:10.3788/AOS201939.0930008

Journals >Acta Optica Sinica >Volume 39 >Issue 9 >Page 0930008 > Article

Acta Optica Sinica
Vol. 39, Issue 9, 0930008 (2019)

Positioning of Characteristic Spectral Peaks Based on Improved Sine Cosine Algorithm

Kun Yu¹, Qingliang Jiao¹, Zilong Liu^1、2、*, Yiqin Jiang², Qiaoxiang Zhang², and Yufang Liu¹

Author Affiliations

¹ College of Physics & Materials Science, Henan Normal University, Xinxiang, Henan 453007, China;

² Division of Optics, National Institute of Metrology, Beijing 100029, China

show less

DOI: 10.3788/AOS201939.0930008 Cite this Article Set citation alerts

Kun Yu, Qingliang Jiao, Zilong Liu, Yiqin Jiang, Qiaoxiang Zhang, Yufang Liu. Positioning of Characteristic Spectral Peaks Based on Improved Sine Cosine Algorithm[J]. Acta Optica Sinica, 2019, 39(9): 0930008 Copy Citation Text

show less

Fig. 1. Measured spectrograms. (a) Ho2O3 filter; (b) Pr-Nd filter

Download full size

Wavelength localization results of spectral characteristic peaks with Gaussian white noise of different signal-to-noise ratios added. (a) Strong peak; (b) weak peak; (c) overlapping peak

Fig. 2. Wavelength localization results of spectral characteristic peaks with Gaussian white noise of different signal-to-noise ratios added. (a) Strong peak; (b) weak peak; (c) overlapping peak

Download full size

No.	Ho₂O₃ filter		Pr-Nd filter
No.	Peak position /nm	Peak type	Peak position /nm		Peak type
1	241.7			Weak		351.1	Strong
2	279.3			Weak		431.2	Weak
3	288.0			Weak		461.4	Weak
4	333.9			Weak		473.5	Weak
5	347.7			Weak		513.4	Strong
6	360.9			Strong		529.8	Strong
7	381.0			Weak		572.7	Weak
8	385.8			Weak		585.7	Strong
9	418.6			Strong		624.4	Weak
10	446.1			Strong		684.8	Weak
11	453.6			Weak		739.2	Strong
12	460.0			Strong		748.3	Weak
13	473.3			Weak		807.7	Strong
14	484.9			Weak
15	536.6			Strong
16	637.8			Weak
17	649.5			Weak
18	662.4			Weak

Table 1. Characteristic peaks of certified reference materials for Pr-Nd filters and Ho2O3 filters

Algorithm	Peak recognition rate
Algorithm	Strong peak /%	Weak peak /%	Overlapping peak /%
Third-order derivative algorithm	100	60^*	37.1^*
Symmetric zero-area algorithm	100	90^*	56.3
SCA	100	85	59.6
Improved SCA	100	95	87.4
*indicates that the incorrect recognition case exists.

Table 2. Peak recognition rate

Algorithm	RMSE /pm
Algorithm	Strong peak	Weak peak	Overlapping peak
Third-order derivative algorithm	1.60	8.19	15.71
Symmetric zero-area algorithm	1.49	7.81	11.33
SCA	1.55	7.88	11.04
Improved SCA	1.53	5.20	9.12

Table 3. RMSE of spectral peak central wavelength

Algorithm	RMSE /%
Algorithm	Strong peak	Weak peak	Overlapping peak
Third-order derivative algorithm	0.12	3.57	3.91
Symmetric zero-area algorithm	0.16	3.93	3.94
SCA	0.11	3.44	3.50
Improved SCA	0.11	3.20	3.37