• Laser & Optoelectronics Progress
  • Vol. 58, Issue 16, 1630004 (2021)
Jinkun Liu1, Chunyu Li1、*, Hang Lü1, Weigang Kong2, Wei Sun1, and Gefei Zhang1
Author Affiliations
  • 1School of Investigation, People's Public Security University of China, Beijing 100038, China
  • 2Institute of Criminal Science and Technology, Zhengzhou Public Security Bureau, Zhengzhou, Henan 450000, China
  • show less
    DOI: 10.3788/LOP202158.1630004 Cite this Article Set citation alerts
    Jinkun Liu, Chunyu Li, Hang Lü, Weigang Kong, Wei Sun, Gefei Zhang. Classification and Recognition of Disposable Masks Based on Raman Spectroscopy and Machine Learning[J]. Laser & Optoelectronics Progress, 2021, 58(16): 1630004 Copy Citation Text show less
    Appearances of masks under different magnifications. (a) 5×; (b) 20×; (c) 50×; (d) 100×
    Fig. 1. Appearances of masks under different magnifications. (a) 5×; (b) 20×; (c) 50×; (d) 100×
    Mask spectra obtained by different wavelength lasers
    Fig. 2. Mask spectra obtained by different wavelength lasers
    Experimental results of reproducibility and uniformity of mask samples. (a) Reproducibility experiment; (b) uniformity experiment
    Fig. 3. Experimental results of reproducibility and uniformity of mask samples. (a) Reproducibility experiment; (b) uniformity experiment
    Raman spectrum processing results of 37 mask samples
    Fig. 4. Raman spectrum processing results of 37 mask samples
    Flow chart of the mask classification model based on Raman spectroscopy
    Fig. 5. Flow chart of the mask classification model based on Raman spectroscopy
    Visualization of the PCA
    Fig. 6. Visualization of the PCA
    Characteristic peaks of the Raman spectrum of the mask. (a) 8#, 10#, 14# , 16#; (b) 1#, 2#, 19#, 31#, 33#; (c) 3#, 12#, 13#, 35#; (d) 21#, 25#, 29#, 30#
    Fig. 7. Characteristic peaks of the Raman spectrum of the mask. (a) 8#, 10#, 14# , 16#; (b) 1#, 2#, 19#, 31#, 33#; (c) 3#, 12#, 13#, 35#; (d) 21#, 25#, 29#, 30#
    Prediction results of the SVM model. (a) Train set; (b) test set
    Fig. 8. Prediction results of the SVM model. (a) Train set; (b) test set
    Prediction results of the Bayesian discriminant analysis model. (a) Train set; (b) test set
    Fig. 9. Prediction results of the Bayesian discriminant analysis model. (a) Train set; (b) test set
    Relationship between iteration times and cross entropy
    Fig. 10. Relationship between iteration times and cross entropy
    LabelBrandManufacturer
    1Xuancheng maskRizhao Nuohuan Protective Equipment Co., Ltd
    2Huian maskZhangjiagang Meibaiqi Trade Co., Ltd
    3ShengWang maskXiantao Siqi Protective Equipment Co., Ltd
    4JiaBoNeng maskGuangdong Dongwan Yian labor protection products Co., Ltd
    33Huian maskSuzhou Lotte Protective Equipment Co., Ltd
    34ZHICHANG maskNanchang of Jiangxi Province
    35Sanbang maskFoshan Nanhai Weijian Sanbang Protective Equipment Technology Co., Ltd
    36Taibang maskYunnan Baiyao Group Co., Ltd
    373M mask3M China Ltd
    Table 1. Sample parameters of the disposable protective masks
    PCEigenvalueVariance /%Cumulative variance /%
    118.8033549.0249.02
    213.8590136.1385.15
    32.361976.1691.31
    41.769374.6195.92
    51.066302.7898.70
    60.134230.3599.05
    70.095340.2599.30
    80.062650.1699.46
    90.033210.0999.55
    100.029420.0899.63
    110.024000.0699.69
    120.021080.0599.74
    130.017520.0599.79
    140.013110.0399.82
    150.011480.0399.85
    160.010010.0399.88
    170.008000.0299.90
    180.007290.0299.92
    190.005250.0199.93
    200.004710.0199.95
    210.003870.0199.96
    220.003410.0199.97
    230.002500.0199.97
    240.002020.0199.98
    250.001680.0099.98
    260.001420.0099.99
    270.001160.0099.99
    280.000900.0099.99
    290.000840.0099.99
    300.000630.0099.99
    310.000600.00100.00
    Table 2. Contribution rate of the PCA characteristic variance
    ClassSample number
    11#、2#、19#
    231#
    333#
    414#
    54#、7#、8#、9#、10#、11#、16#、17#、18#、20#、22#、23#、26#、27#、29#、34#、37#
    63#、5#、6#、12#、13#、15#、21#、24#、25#、28#、30#、32#、35#、36#
    Table 3. Mask classification results based on PCA and Raman spectra
    Kernel function typeTraining set accuracy /%Test set accuracy /%Run time /s
    Linear50.028.65
    Polynomial46.542.910
    RBF93.3100.020
    Sigmoid96.785.015
    Table 4. Influence of different kernel functions on the SVM model
    FunctionEigenvalueVariance /%Cumulative variance /%Canonical correlationFunction testWilks’lambdaSig
    115064.48766.266.211--50.000.00
    26222.54027.493.612--50.000.00
    31303.9515.799.313--50.000.00
    4126.6700.699.90.9964--50.000.00
    524.2170.11000.98050.040.00
    Table 5. Bayesian discriminant function of mask samples
    Jinkun Liu, Chunyu Li, Hang Lü, Weigang Kong, Wei Sun, Gefei Zhang. Classification and Recognition of Disposable Masks Based on Raman Spectroscopy and Machine Learning[J]. Laser & Optoelectronics Progress, 2021, 58(16): 1630004
    Download Citation