• Acta Optica Sinica
  • Vol. 42, Issue 22, 2230002 (2022)
Hailong Zhao1, Shu Gan1、2、*, Xiping Yuan2、3, Lin Hu1, Shuai Liu1, and Junjie Wang1
Author Affiliations
  • 1Faculty of Land and Resources Engineering, Kunming University of Science and Technology, Kunming 650093, Yunnan , China
  • 2Yunnan Institute of Engineering Research and Application of Plateau Mountain Spatial Information Surveying and Mapping Technology, Kunming 650093, Yunnan , China
  • 3West Yunnan University of Applied Sciences, Dali671000, Yunnan , China
  • show less
    DOI: 10.3788/AOS202242.2230002 Cite this Article Set citation alerts
    Hailong Zhao, Shu Gan, Xiping Yuan, Lin Hu, Shuai Liu, Junjie Wang. Inversion of Soil Iron Oxide Based on Multi-Scale Continuous Wavelet Decomposition[J]. Acta Optica Sinica, 2022, 42(22): 2230002 Copy Citation Text show less
    Distribution map of soil sampling points
    Fig. 1. Distribution map of soil sampling points
    Original spectral reflectance curves of all soil samples
    Fig. 2. Original spectral reflectance curves of all soil samples
    Correlation coefficient between spectrum and iron oxide content
    Fig. 3. Correlation coefficient between spectrum and iron oxide content
    Thermodynamic diagram of wavelet coefficient and coefficient of determination of iron oxide content
    Fig. 4. Thermodynamic diagram of wavelet coefficient and coefficient of determination of iron oxide content
    Characteristic wavelengths selected by CC-CARS algorithm
    Fig. 5. Characteristic wavelengths selected by CC-CARS algorithm
    Scatter plot of measured and predicted values of soil iron oxide content under different models. (a) L4-CC-CARS-SVR; (b) OS-CC-CARS-SVR; (c) FD-CC-CARS-SVR; (d) RL-CC-CARS-SVR
    Fig. 6. Scatter plot of measured and predicted values of soil iron oxide content under different models. (a) L4-CC-CARS-SVR; (b) OS-CC-CARS-SVR; (c) FD-CC-CARS-SVR; (d) RL-CC-CARS-SVR
    Sample classificationSample number

    Maximum /

    (g·kg-1

    Minimum /

    (g·kg-1

    Mean /

    (g·kg-1

    Standard deviation /

    (g·kg-1

    Variable coefficient /%
    Total set13566.97818.29341.20111.69828.393
    Calibration set9564.80823.31142.14110.73625.476
    Validation set4066.97818.29338.96913.60534.912
    Red soil6866.97823.29746.83811.80425.016
    Purple soil4862.36523.31135.8347.84021.880
    Yellow brown soil1954.35618.29334.5839.55928.400
    Table 1. Statistical characteristics of iron oxide content
    Transform spectrumWavelet decomposition scaleSignificant band numberMaximum correlation coefficient
    CWTL192-0.590
    L2171-0.593
    L33950.606
    L4663-0.602
    L511360.603
    L61056-0.604
    L713200.527
    L81357-0.511
    L91273-0.548
    L101447-0.523
    OS2051-0.589
    FD1252-0.548
    RL20510.606
    Table 2. Correlation analysis of transform spectrum and iron oxide content
    Transform spectrumCalibration setValidation set
    R2ERMSE /(g·kg-1R2ERMSE /(g·kg-1RPIQ
    L10.7834.9690.16812.2511.654
    L20.6915.9320.37010.6571.901
    L30.8254.4640.5908.6002.356
    L40.7605.2360.6637.7982.598
    L50.6226.5650.6378.0952.503
    L60.5487.1830.4509.9672.034
    L70.7695.1330.6188.2982.442
    L80.5117.4630.4989.5172.129
    L90.3218.8010.10212.7261.592
    L100.3409.5930.34211.5481.755
    OS0.6786.0620.6168.3282.433
    FD0.6476.3430.6388.0762.509
    RL0.7445.4020.6378.0922.504
    Table 3. Results of soil iron oxide inversion model
    Hailong Zhao, Shu Gan, Xiping Yuan, Lin Hu, Shuai Liu, Junjie Wang. Inversion of Soil Iron Oxide Based on Multi-Scale Continuous Wavelet Decomposition[J]. Acta Optica Sinica, 2022, 42(22): 2230002
    Download Citation