• Spectroscopy and Spectral Analysis
  • Vol. 41, Issue 11, 3385 (2021)
Yong HAO1、1;, Jiao-jun DU1、1;, Shu-min ZHANG2、2;, and Qi-ming WANG1、1;
Author Affiliations
  • 11. School of Mechatronics & Vehicle Engineering, East China Jiaotong University, Nanchang 330013, China
  • 22. Nanchang Customs Technology Center, Nanchang 330013, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2021)11-3385-07 Cite this Article
    Yong HAO, Jiao-jun DU, Shu-min ZHANG, Qi-ming WANG. Research on Construction of Visible-Near Infrared Spectroscopy Analysis Model for Soluble Solid Content in Different Colors of Jujube[J]. Spectroscopy and Spectral Analysis, 2021, 41(11): 3385 Copy Citation Text show less
    Samples of Dali jujube with different colours
    Fig. 1. Samples of Dali jujube with different colours
    Schematic diagram of spectrum acquisition device
    Fig. 2. Schematic diagram of spectrum acquisition device
    Average spectra of jujube samples
    Fig. 3. Average spectra of jujube samples
    Original Vis-NIR spectra of Jujube samples and spectra after pretreatment with SNV(a): Original Vis-NIR spectra; (b): Spectra after pretreatment with SNV
    Fig. 4. Original Vis-NIR spectra of Jujube samples and spectra after pretreatment with SNV
    (a): Original Vis-NIR spectra; (b): Spectra after pretreatment with SNV
    RMSEP variation diagram of PLS models for jujube SSC established by samples with different colours for modeling sets
    Fig. 5. RMSEP variation diagram of PLS models for jujube SSC established by samples with different colours for modeling sets
    Distribution of selected variables by MCUVE method
    Fig. 6. Distribution of selected variables by MCUVE method
    Variations of RMSECV and RMSEP of mixed jujube model with PLS factors
    Fig. 7. Variations of RMSECV and RMSEP of mixed jujube model with PLS factors
    Correlation of SSC measured value and model predicted value
    Fig. 8. Correlation of SSC measured value and model predicted value
    Type of colorCalibration set1ncSSC(°Brix)Prediction set2nvSSC(°Brix)
    MJMJc3010.93~15.67MJv1510.93~15.67
    GJGJc3011.00~17.10GJv1511.50~16.90
    RJRJc3010.60~19.20RJv1511.10~18.30
    MJ-GJMJc-GJc6010.93~17.10MJv-GJv3010.93~16.90
    MJ-GJ-RJMJc-GJc-RJc9010.60~19.20MJv-GJv-RJv4510.93~18.30
    Table 1. Statistics of SSC contents in jujube samples
    SamplesPretreatment
    methods
    FactorRMSECVRc
    MJcRAW80.8980.700
    NWS140.8430.756
    CWD120.9480.672
    MSC100.9580.652
    SNV100.9690.643
    NWS-MSC110.8960.722
    GJcRAW190.8250.854
    NWS100.7180.893
    CWD90.6590.910
    MSC90.7740.873
    SNV200.7390.884
    NWS-MSC100.7170.897
    RJcRAW71.2680.816
    NWS71.2110.835
    CWD61.3370.796
    MSC81.3760.784
    SNV81.3250.799
    NWS-MSC81.3400.801
    MJc-GJcRAW61.3620.496
    NWS101.3290.578
    CWD81.2451.606
    MSC81.1240.676
    SNV81.1320.671
    NWS-MSC111.1080.698
    MJc-GJc-RJcRAW101.3510.786
    NWS111.3180.740
    CWD81.2550.757
    MSC81.1870.785
    SNV81.1560.796
    NWS-MSC101.1680.793
    Table 2. Modelling results of PLS models for jujube SSC with different pretreatment methods
    Samples for modelingPretreatment methodsFactorRMSECVRcPrediction setRMSEPRp
    MJv1.3140.816
    MJcNWS140.8430.752GJv2.1120.657
    RJv2.7920.405
    MJv1.7690.611
    GJcCWD90.6590.910GJv0.8750.822
    RJv1.2760.812
    MJv1.5170.736
    RJcNWS81.2110.835GJv1.5780.831
    RJv0.9600.923
    Table 3. Predicted results of PLS models for SSC of single color jujube
    Samples for modelingPretreatment methodsFactorRMSECVRcPrediction set compositionRMSEPRp
    MJc-GJcNWS-MSC111.1080.698MJv0.9800.724
    GJv0.9830.822
    RJv1.9280.597
    MJv-GJv-RJv1.3780.704
    MJc-GJc-RJcSNV81.1580.796MJv1.0770.668
    GJv0.8810.861
    RJv1.1400.841
    MJv-GJv-RJv0.9460.864
    Table 4. Predicted results of PLS models for SSC of mixed color jujube samples
    Calibration setSelected the number of variablesFactorRMSECVRcPrediction setRMSEPRp
    MJc2 040130.8500.748MJv1.0700.824
    GJc34090.5150.947GJv0.4430.958
    RJc180110.7250.943RJv0.9800.947
    MJc-GJc160140.9030.812MJv0.7970.848
    300140.8080.851GJv0.6050.933
    MJc-GJc-RJc200100.8860.884MJv-GJv-RJv0.7210.922
    Table 5. Analysis results of the models with modelling variables optimization
    Yong HAO, Jiao-jun DU, Shu-min ZHANG, Qi-ming WANG. Research on Construction of Visible-Near Infrared Spectroscopy Analysis Model for Soluble Solid Content in Different Colors of Jujube[J]. Spectroscopy and Spectral Analysis, 2021, 41(11): 3385
    Download Citation