Acta Photonica Sinica, Volume. 52, Issue 11, 1110002(2023)

Multi-scale Remote Sensing Image Classification Based on Weighted Feature Fusion

Yinzhu CHENG1...2, Song LIU1,2, Nan WANG1,2, Yuetian SHI1,2 and Geng ZHANG1,* |Show fewer author(s)
Author Affiliations
  • 1Xi'an Institute of Optics and Precision Mechanics,Chinese Academy of Sciences,Xi'an 710119,China
  • 2University of Chinese Academy of Sciences,Beijing 100049,China
  • show less
    Figures & Tables(14)
    multi-branch 3D CNN framework for HSIC
    Rotational data augmentation for HSIC
    Schematic diagram of 3D convolution layer
    The network structure diagram of the three branches in the training phase
    Classification results for the IP dataset using 10% of the available labeled data
    Classification results for the UP dataset using 10% of the available labeled data
    Classification results for the SV dataset using 10% of the available labeled data
    The relationship between the proportion of the available labeled data and classification results
    Airborne hyperspectral dataset and its classification result map
    • Table 1. Sample distribution of hyperspectral remote sensing dataset

      View table
      View in Article

      Table 1. Sample distribution of hyperspectral remote sensing dataset

      Indian PinesPavia UniversitySalinas
      ClassSamplesClassSamplesClassSamples
      Asphalt46Asphalt6 631Brocoli-green-weeds-12 009
      Corn-notill1 428Meadows18 649Brocoli-green-weeds-23 726
      Corn-mintill830Gravel2 099Fallow1 976
      Corn237Trees-painted3 064Fallow-rough-plow1 394
      Grass-pasture483Metal-sheets1 345Fallow-smooth2 678
      Grass-trees730Bare soil5 029Stubble3 959
      Grass-pasture-mowed28Bitumen1 330Celery3 579
      Hay-windrowed478Blocking-bricks3 682Grapes-untrained11 271
      Oats20Shadows947Soil-vinyard-develop6 203
      Soybean-notill972Corn-senesced-green-weeds3 278
      Soybean-mintill2 455Lettuce-romaine-4wk1 068
      Soybean-clean593Lettuce-romaine-5wk1 927
      Wheat205Lettuce-romaine-6wk916
      Woods1 265Lettuce-romaine-7wk1 070
      Building-grass-Trees-drives386Vinyard-untrained7 268
      Stone-steel-towers93Vinyard-vertical-trellis1 807
      Total10 249Total42 776Total54 129
    • Table 2. Classification results for the IP dataset using 10% of the available labeled data

      View table
      View in Article

      Table 2. Classification results for the IP dataset using 10% of the available labeled data

      ClassSVM/%LSTM/%GRU/%CDCNN/%DBDA/%DBMA/%Ours/%
      Asphalt73.1746.3451.2151.7560.0093.8295.37
      Corn-notill78.3781.7878.5277.6495.1097.0397.66
      Corn-mintill70.5571.3561.9877.1396.5797.1199.95
      Corn72.3061.0372.3078.9298.9898.5899.23
      Grass-pasture90.8081.1488.0491.8696.8397.0394.93
      Grass-trees90.1195.8989.6495.4597.6098.6699.26
      Grass-pasture-mowed68.0080.0068.0054.2058.8280.0197.40
      Hay-windrowed97.6799.5397.6790.4696.7199.90100.00
      Oats61.1177.7750.0063.2740.0081.5497.22
      Soybean-notill71.3175.8873.8281.3488.2495.3098.37
      Soybean-mintill83.3074.7878.8179.9796.6694.0698.73
      Soybean-clean73.7830.1480.7164.9890.7295.4998.56
      Wheat95.6897.8398.3797.89100.0098.4697.89
      Woods95.4396.1394.2993.9996.7297.3699.73
      Building-grass-trees-drivers63.1145.5358.5083.3196.5693.6498.17
      Stone-steel-towers88.1088.0991.6692.3993.2396.8995.89
      OA82.0577.6580.4381.3595.0796.0298.60
      AA79.5575.2077.1079.6687.6794.6898.02
      Kappa79.4774.5177.6678.6794.3895.5598.41
    • Table 3. Classification results for the UP dataset using 10% of the available labeled data

      View table
      View in Article

      Table 3. Classification results for the UP dataset using 10% of the available labeled data

      ClassSVM/%LSTM/%GRU/%CDCNN/%DBDA/%DBMA/%Ours/%
      Asphalt93.3793.1292.6995.4899.0499.6099.91
      Meadows97.8197.5497.1198.7799.8699.86100.00
      Gravel78.0075.9179.1795.7396.1999.4599.08
      Trees-painted95.3292.9395.3098.8599.0199.0499.28
      Metal-sheets99.4299.6399.7599.8399.9399.6199.88
      Bare soil90.3786.8688.1795.2599.9399.90100.00
      Bitumen87.9783.9687.0593.30100.00100.0099.83
      Blocking-bricks88.6187.2485.9589.3798.2797.6499.66
      Shadows99.9499.8899.7199.0799.0899.4299.53
      OA94.0993.0193.2396.5799.3399.5399.83
      AA92.3190.7991.6596.1899.0399.3999.68
      Kappa92.1690.7191.0295.4699.1199.3899.78
    • Table 4. Classification results for the SV dataset using 10% of the available labeled data

      View table
      View in Article

      Table 4. Classification results for the SV dataset using 10% of the available labeled data

      ClassSVM/%LSTM/%GRU/%CDCNN/%DBDA/%DBMA/%Ours/%
      Brocoli_g_w_199.5699.6999.8679.94100.0099.23100.00
      Brocoli_g_w_299.9499.6399.8892.8899.9899.99100.00
      Fallow99.4999.6999.7198.6099.9298.52100.00
      Fallow_r_p99.5299.5699.6098.7499.3598.8499.92
      Fallow_s99.5998.2899.0489.0099.4499.5599.96
      Stubble99.9299.8499.8999.8799.92100.00100.00
      Celery99.6699.7899.7699.7299.9299.94100.00
      Grapes_u90.2087.5687.1185.7896.0999.8099.95
      Soil_v_d99.9199.0299.6799.7999.9499.84100.00
      Corn_s_g_w97.6995.7497.6797.6299.8098.8499.97
      Lettuce_r_4wk99.0698.0298.5993.85100.0099.19100.00
      Lettuce_r_5wk99.7796.8699.4299.4599.9799.73100.00
      Lettuce_r_6wk99.3998.2499.2198.79100.0099.27100.00
      Lettuce_r_7wk97.9296.6297.8297.7699.7799.7499.83
      Vinyard_u71.2482.1582.5583.5697.8694.5999.96
      Vinyard_v_t99.2098.8398.8693.0586.7799.9699.80
      OA93.7494.1894.5591.9997.4598.8999.97
      AA97.0096.8597.4294.2798.6799.1899.96
      Kappa93.0293.5393.9391.0997.1698.7799.96
    • Table 5. The overall classification accuracy of each algorithm when using 2% labeled samples for training

      View table
      View in Article

      Table 5. The overall classification accuracy of each algorithm when using 2% labeled samples for training

      AlgorithmIP/%UP/%SV/%
      DBDA87.2997.8196.28
      DBMA78.3597.8598.14
      OURS89.6899.2999.86
    Tools

    Get Citation

    Copy Citation Text

    Yinzhu CHENG, Song LIU, Nan WANG, Yuetian SHI, Geng ZHANG. Multi-scale Remote Sensing Image Classification Based on Weighted Feature Fusion[J]. Acta Photonica Sinica, 2023, 52(11): 1110002

    Download Citation

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

    Category:

    Received: Apr. 4, 2023

    Accepted: May. 22, 2023

    Published Online: Dec. 22, 2023

    The Author Email: ZHANG Geng (gzhang@opt.ac.cn)

    DOI:10.3788/gzxb20235211.1110002

    Topics