Characteristic Wavelength Optimization Based on Random Frog Algorithm

Jiming Sa; He Jiang; Kaiwen Xie; Hanwen Gu; Yijie Luo; Zhushanying Zhang

doi:10.3788/AOS202141.1530001

Acta Optica Sinica, Volume. 41, Issue 15, 1530001(2021)

Characteristic Wavelength Optimization Based on Random Frog Algorithm

Jiming Sa^1,2,3, He Jiang^1、*, Kaiwen Xie¹, Hanwen Gu¹, Yijie Luo², and Zhushanying Zhang^3,4,5、**

Author Affiliations

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

²Hubei Key Laboratory of Broadband Wireless Communication and Sensor Networks, Wuhan, Hubei 430070, China

³Hubei Key Laboratory of Medical Information Analysis and Tumor Diagnosis & Treatment, Wuhan, Hubei 430074, China;

⁴College of Biomedical Engineering, South-Central University for Nationalities, Wuhan, Hubei 430074, China

⁵Key Laboratory of Cognitive Science (South-Central University for Nationalities), State Ethnic Affairs Commission, Wuhan, Hubei 430074, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(16)

Fig. 1. Spectra of hemoglobin phantom solution sample

Download full size

Fig. 2. Spectra of blood sample

Download full size

Fig. 3. Process of WRF algorithm

Download full size

Fig. 4. RMSECV and number of principal factors corresponding to different intervals for phantom solution sample

Download full size

Fig. 5. Characteristic wavelength distribution based on SPA for phantom solution sample

Download full size

Fig. 6. Selection process of characteristic wavelength based on CARS for phantom solution sample. (a) Number of wavelengths; (b) RMSECV; (c) return path

Download full size

Fig. 7. Optimization results of characteristic wavelengths based on WRF for phantom solution sample. (a) The RMSECV value of the merge intervals; (b) the wavelength distribution

Download full size

Fig. 8. RMSECV and number of principal factors corresponding to different intervals for blood sample

Download full size

Fig. 9. Characteristic wavelength distribution based on SPA for blood sample

Download full size

Fig. 10. Selection process of characteristic wavelength based on CARS for blood sample. (a) Number of wavelengths; (b) RMSECV; (c) return path

Download full size

Fig. 11. Optimization results of characteristic wavelengths based on WRF for blood sample. (a) The RMSECV value of the merge intervals; (b) the wavelength distribution

Download full size

Table 1. Statistics of hemoglobin content

View table

Table 1. Statistics of hemoglobin content

Sample	Sample set	Total numberof samples	Mass concentration of hemoglobin /(g·L^-1)
Sample	Sample set	Total numberof samples	Maximum	Minimum	Average	Standarddeviation
	Entire set	146	150	5	77.50	42.29
Phantomsolution sample	Calibration set	110	150	5	77.34	42.53
	Validation set	36	148	8	78.00	42.14
	Entire set	185	173	103	137.08	16.49
Bloodsample	Calibration set	139	173	103	137.95	16.40
	Validation set	46	173	106	134.46	16.66

Table 2. Parameter optimization results of successive projections algorithm
View table
Table 2. Parameter optimization results of successive projections algorithm
Maximum 10 20 30 40 50 60 70
Number of wavelengths 2 2 4 5 5 5 5
RMSECV 4.3313 4.3313 2.2511 2.1494 2.1494 2.1494 2.1494

Table 3. Optimization results of characteristic wavelength of phantom solution sample
View table
Table 3. Optimization results of characteristic wavelength of phantom solution sample
Optimizationalgorithm Numberof principalcomponents RMSECV RMSEP R_p
Full 9 3.3454 4.4792 0.9948
SiPLS 13 2.8266 3.8059 0.9960
SPA 6 2.1494 3.4661 0.9965
CARS 5 2.1420 2.1282 0.9985
WRF 4 2.2962 2.7742 0.9984

Table 4. Optimization results of characteristic wavelength of blood sample
View table
Table 4. Optimization results of characteristic wavelength of blood sample
Optimizationalgorithm Numberof principalcomponents RMSECV RMSEP R_p
Full 14 3.8142 5.3758 0.9421
SiPLS 10 2.3614 3.3581 0.9763
SPA 15 3.7454 2.4686 0.9707
CARS 10 5.1743 5.2252 0.9470
WRF 9 2.7972 2.3205 0.9828

Table 5. Result comparison of different algorithms
View table
Table 5. Result comparison of different algorithms
Algorithm Characteristicwavelength number RMSEP R_p
SG+PLS 700 5.45 0.940
RF+PLS 34 3.35 0.974
WRF+PLS 31 2.79 0.983

Tools

Get Citation

Copy Citation Text

Jiming Sa, He Jiang, Kaiwen Xie, Hanwen Gu, Yijie Luo, Zhushanying Zhang. Characteristic Wavelength Optimization Based on Random Frog Algorithm[J]. Acta Optica Sinica, 2021, 41(15): 1530001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites