Spectroscopy and Spectral Analysis, Volume. 42, Issue 7, 2247(2022)

Similar Wood Species Classification Within Pterocarpus Genus Using Feature Fusion

Cheng-kun WANG2、*, Peng ZHAO1、1; 2; *;, and Xiang-hua LI2、2;
Author Affiliations
  • 11. College of Computer Science and Electronics, Guangxi University of Science and Technology, Liuzhou 545006, China
  • 22. College of Information and Computer Engineering, Northeast Forestry University, Harbin 150040, China
  • show less
    Figures & Tables(15)
    Wood feature acquisition platforms(a): Spectrum acquisition; (b): RGB image acquisition
    Images of Pterocarpus section(a): Pterocarpus macrocarpus; (b): Pterocarpus erinaceus; (c): Pterocarpus antunesii;(d): Pterocarpus soyauxii; (e): Pterocarpus tinctorius
    Original spectra and SNV corrected spectra(a): Original spectra; (b): SNV corrected spectra
    Feature dimension and accuracy
    Identification of test set samples
    Average spectral curves of cross sections of 30 wood species(a): The first 15 tree species in Table 7;(b): The last 15 tree species in Table 7
    Transverse section of 30 wood species (the number of each illustrations corresponding to Table 7)
    • Table 1. Sample data

      View table
      View in Article

      Table 1. Sample data

      序号中文拉丁文
      1大果紫檀Pterocarpus macrocarpus
      2刺猬紫檀Pterocarpus erinaceus
      3安氏紫檀Pterocarpus antunesii
      4非洲紫檀Pterocarpus soyauxii
      5赞比亚紫檀Pterocarpus tinctorius
    • Table 2. The highest accuracies under different dimension reduction methods

      View table
      View in Article

      Table 2. The highest accuracies under different dimension reduction methods

      切面横切面弦切面径切面
      方法维度正确率/%维度正确率/%维度正确率/%
      PCA2894.402690.403093.20
      KPCA7689.205686.406088.40
      Laplacian3282.802878.403682.80
      SPA选择波段/nm正确率/%
      横切面376.64, 378.97, 386.31, 495.03, 630.43, 695.46, 806.52, 1 010.28, 1 017.95, 1 025.29, 1 026.2992.80
      弦切面378.30, 408.65, 529.04, 600.41, 713.14, 849.208, 1 019.29, 1 021.96, 1 025.29, 1 025.9691.60
      径切面377.64, 379.97, 586.74, 745.15, 937.91, 995.28, 1 016.62, 1 025.29, 1 025.96, 1 026.2993.60
    • Table 3. Accuracies of wood species using textures features (%)

      View table
      View in Article

      Table 3. Accuracies of wood species using textures features (%)

      方法横切面弦切面径切面
      GLCM67.6063.2064.80
      LBP80.0077.6074.00
      I-BGLAM75.6072.4075.60
      MFS62.0068.0063.20
    • Table 4. Accuracies of “concat” and “sum” fusion schemes (%)

      View table
      View in Article

      Table 4. Accuracies of “concat” and “sum” fusion schemes (%)

      融合策略concatsum
      方法PCAKPCALaplacianSPAPCAKPCALaplacianSPA
      横切面
      GLCM93.6082.8082.4097.6092.4082.8081.6096.00
      I-BGLAM99.2088.4083.2098.4098.8090.0082.0097.20
      MFS98.0086.8086.4099.2092.8083.2086.0096.40
      LBP96.8087.6088.0098.0093.6089.2090.0095.20
      弦切面
      GLCM96.0092.8092.4098.8094.8092.4090.4097.20
      I-BGLAM99.2091.2084.0098.8099.2088.8082.0098.40
      MFS90.8080.8078.8098.4089.2078.4078.4092.00
      LBP93.6089.6083.2098.4092.0088.8084.8095.60
      径切面
      GLCM97.6091.2090.8098.4098.0092.8090.8098.40
      I-BGLAM98.8089.6088.0099.2098.8089.2086.8099.20
      MFS92.4082.4080.8096.4090.4081.2082.4092.80
      LBP99.2088.0086.8098.8096.0086.4085.2097.20
    • Table 5. Comparison of accuracies between other wood recognition methods and the method presented in this paper

      View table
      View in Article

      Table 5. Comparison of accuracies between other wood recognition methods and the method presented in this paper

      方法正确率/%
      CNN80.00
      颜色矩74.40
      SPPD+I-BGLAM77.60
      Fuzzy+SPPD+I-BGLAM73.60
      GA56.00
      GA+KDA57.60
      本方法99.20
    • Table 6. Extraction time of single sample features

      View table
      View in Article

      Table 6. Extraction time of single sample features

      方法时间/s
      纹理GLCM0.017
      I-BGLAM0.032
      MFS1.32
      LBP0.033
      光谱PCA0.002 5
      KPCA0.000 14
      Laplacian0.000 71
      SPA0.72
      融合CCA0.002 2
    • Table 7. Details of 30 wood species samples

      View table
      View in Article

      Table 7. Details of 30 wood species samples

      序号名称拉丁语
      1海棠木红厚壳属Calophyllum inophyllum
      2香樟木樟木属Cinnamomum camphora
      3大非洲楝非洲楝属Entandrophragma candoLaplaciani
      4美洲白蜡木白蜡树属Fraxinus chinensis
      5水曲柳白蜡树属Fraxinus mandshurica
      6古夷苏木古夷苏木属Guibourtia demeusei
      7双柱苏木古夷苏木属Guibourtia ehie
      8帕利印茄印茄属Intsia bijuga
      9黑核桃核桃树Juglans nigra
      10落叶松落叶松属Larix gmelinii
      11黑芯木莲木莲属Magnolia fordiana
      12非洲崖豆木崖豆藤属MiLaplacianttia laurentii
      13云杉云杉属Picea asperata
      14辐射松松属Pinups radiata
      15红松松属Pinus koraiensis
      16马尾松松属Pinus massoniana
      17樟子松松属Pinus sylvestris
      18番龙眼番龙眼属Pometia pinnata
      19花旗松木黄杉属Pseudotsuga menziesii
      20柞木麻栎属Quercus mongolica
      21麻栎麻栎属Quercus acutissima
      22刺槐刺槐Robinia pseudoacacia
      23无齿婆罗双婆罗双属Shorea contorta
      24平滑娑罗双婆罗双属Shorea laevis
      25槐树槐树属Sophora japonica
      26桃花芯桃花心木属Swietenia mahagoni
      27缅甸柚木柚木属Tectona grandis
      28榄仁木榄仁树属Terminalia cattapa
      29榆树榆树属Ulmus glabra
      30油桐油桐属Vernicia fordii
    • Table 8. Classification accuracy of 35 tree species data

      View table
      View in Article

      Table 8. Classification accuracy of 35 tree species data

      方法正确率/%
      纹理特征I-BGLAM
      LBP
      67.14
      65.89
      光谱特征PCA
      SPA
      91.89
      93.09
      融合特征PCA+I-BGLAM
      PCA+LBP
      SPA+I-BGLAM
      SPA+LBP
      97.77
      95.20
      96.57
      98.29
    Tools

    Get Citation

    Copy Citation Text

    Cheng-kun WANG, Peng ZHAO, Xiang-hua LI. Similar Wood Species Classification Within Pterocarpus Genus Using Feature Fusion[J]. Spectroscopy and Spectral Analysis, 2022, 42(7): 2247

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Orginal Article

    Received: Jun. 7, 2021

    Accepted: --

    Published Online: Nov. 16, 2022

    The Author Email: Cheng-kun WANG (402686820@qq.com)

    DOI:10.3964/j.issn.1000-0593(2022)07-2247-08

    Topics