• Spectroscopy and Spectral Analysis
  • Vol. 42, Issue 4, 1229 (2022)
Tian-liang ZHANG*, Dong-xing ZHANG, Tao CUI, Li YANG*;, Chun-ji XIE, Zhao-hui DU, and Xiang-jun ZHONG
Author Affiliations
  • Key Laboratory of Soil-Machine-Plant System Technology of Ministry of Agriculture, College of Engineering, China Agricultural University, Beijing 100083, China
  • show less
    DOI: 10.3964/j.issn.1000-0593(2022)04-1229-06 Cite this Article
    Tian-liang ZHANG, Dong-xing ZHANG, Tao CUI, Li YANG, Chun-ji XIE, Zhao-hui DU, Xiang-jun ZHONG. Identification of Early Lodging Resistance of Maize by Hyperspectral Imaging Technology[J]. Spectroscopy and Spectral Analysis, 2022, 42(4): 1229 Copy Citation Text show less
    Spectral image correction and spectral extraction method(a): Color image of leaves; (b): k-means cluster result; (c): Leaves segmentation result
    Fig. 1. Spectral image correction and spectral extraction method
    (a): Color image of leaves; (b): k-means cluster result; (c): Leaves segmentation result
    Average spectral curves of lodging-resistant varieties and lodging varieties
    Fig. 2. Average spectral curves of lodging-resistant varieties and lodging varieties
    Spectral characteristics selected by the SPA algorithm at various densities
    Fig. 3. Spectral characteristics selected by the SPA algorithm at various densities
    Parameter optimization effects of SPA-SVM model at various densities
    Fig. 4. Parameter optimization effects of SPA-SVM model at various densities
    Parameter optimization effects of PCA-SVM model at various densities
    Fig. 4. Parameter optimization effects of PCA-SVM model at various densities
    ROC curve of model prediction effect at each density
    Fig. 5. ROC curve of model prediction effect at each density
    项目5 000株·亩-17 000株·亩-19 000株·亩-1
    抗倒样本不抗倒样本总计抗倒样本不抗倒样本总计抗倒样本不抗倒样本总计
    训练集959118696961929495189
    测试集303060303060303060
    总计125121246126126252124125249
    Table 1. The division of maize sample data sets
    主成分5 000 株·亩-17 000 株·亩-19 000 株·亩-1
    贡献率
    /%
    累计
    贡献率
    /%
    贡献率
    /%
    累计
    贡献率
    /%
    贡献率
    /%
    累计
    贡献率
    /%
    PC149.2849.2845.5945.5948.3148.31
    PC234.3083.5832.1077.6936.0684.37
    PC38.7792.3512.4290.117.5891.95
    PC43.0795.423.1693.273.4095.35
    PC51.9297.342.5295.791.7697.11
    PC61.1398.471.1596.940.8797.78
    PC70.5098.971.0998.030.6198.59
    PC80.2199.180.5698.590.4199.00
    PC90.1999.370.3998.980.2499.24
    PC100.1799.540.2599.230.1899.42
    Table 2. Contribution rate of principal component at each density
    项目5 000株·亩-17 000株·亩-19 000株·亩-1
    PCA-SVMSPA-SVMPCA-SVMSPA-SVMPCA-SVMSPA-SVM
    惩罚参数C2921721121923211
    核参数γ2-112-132-112-152-72-7
    交叉验证正确率/%79.0310083.3397.4076.1998.94
    测试集分类正确率/%78.3383.3383.3398.3371.6795
    Table 3. The selected model parameters at each density
    Tian-liang ZHANG, Dong-xing ZHANG, Tao CUI, Li YANG, Chun-ji XIE, Zhao-hui DU, Xiang-jun ZHONG. Identification of Early Lodging Resistance of Maize by Hyperspectral Imaging Technology[J]. Spectroscopy and Spectral Analysis, 2022, 42(4): 1229
    Download Citation