• Spectroscopy and Spectral Analysis
  • Vol. 42, Issue 9, 2931 (2022)
Jiang-sheng GUI1、*, Jie HE1、1;, and Xia-ping FU2、2;
Author Affiliations
  • 11. School of Information Science and Technology, Zhejiang Sci-Tech University, Hangzhou 310018, China
  • 22. Faculty of Mechanical Engineering & Automation, Zhejiang Sci-Tech University, Hangzhou 310018, China;
  • show less
    DOI: 10.3964/j.issn.1000-0593(2022)09-2931-04 Cite this Article
    Jiang-sheng GUI, Jie HE, Xia-ping FU. Hyperspectral Detection of Soybean Heart-Eating Insect Pests Based on Image Retrieval[J]. Spectroscopy and Spectral Analysis, 2022, 42(9): 2931 Copy Citation Text show less
    Hyperspectral imaging system
    Fig. 1. Hyperspectral imaging system
    Hyperspectral images of soybean samples(a): Normal soybean; (b): Soybean with egg;(c): Soybean with larvae; (d): Gnawed soybean
    Fig. 2. Hyperspectral images of soybean samples
    (a): Normal soybean; (b): Soybean with egg;(c): Soybean with larvae; (d): Gnawed soybean
    小样本学习模型特征提取模型学习率5-shot准确率/%
    MAMLResnet180.0160.05±0.50
    MNResnet180.0177.65±0.50
    3D-RNResnet180.0182.50±1.50
    Table 1. Detection results of different classification models in 4-way 5-shot
    损失函数预训练模型学习率准确率/%
    交叉熵损失Resnet18-K
    Resnet18-KM
    0.0175.0±1.00
    74.5±1.00
    DSHSD[10]Resnet18-K
    Resnet18-KM
    0.0181.0±1.00
    81.5±1.50
    DH[7]Resnet18-K
    Resnet18-KM
    0.0178.5±1.00
    79.5±1.00
    DSH[8]Resnet18-K
    Resnet18-KM
    0.0180.5±0.50
    81.5±1.00
    CSQ[11]Resnet18-K
    Resnet18-KM
    0.0179.0±1.00
    80.0±1.00
    DCH[6]Resnet18-K
    Resnet18-KM
    0.0185.5±1.00
    86.0±1.00
    Table 2. Retrieval performance under different loss functions
    Jiang-sheng GUI, Jie HE, Xia-ping FU. Hyperspectral Detection of Soybean Heart-Eating Insect Pests Based on Image Retrieval[J]. Spectroscopy and Spectral Analysis, 2022, 42(9): 2931
    Download Citation