• Laser & Optoelectronics Progress
  • Vol. 57, Issue 20, 202803 (2020)
Qing Fu1、2、3, Chen Guo1、2、*, and Wenlang Luo1、2
Author Affiliations
  • 1School of Electronics and Information Engineering, Jinggangshan University, Ji'an, Jiangxi 343009, China
  • 2Jiangxi Engineering Laboratory of IoT Technologies for Crop Growth, Ji'an, Jiangxi 343009, China
  • 3College of Surveying and Geo-Informatics, Tongji University, Shanghai 200092, China
  • show less
    DOI: 10.3788/LOP57.202803 Cite this Article Set citation alerts
    Qing Fu, Chen Guo, Wenlang Luo. A Hyperspectral Image Classification Method Based on Spectral-Spatial Features[J]. Laser & Optoelectronics Progress, 2020, 57(20): 202803 Copy Citation Text show less
    Hyperspectral image classification based on Log-Gabor filtering and CNN
    Fig. 1. Hyperspectral image classification based on Log-Gabor filtering and CNN
    Training loss curve and accuracy curve in the Pavia University dataset. (a) Training loss curve; (b) training accuracy curve
    Fig. 2. Training loss curve and accuracy curve in the Pavia University dataset. (a) Training loss curve; (b) training accuracy curve
    Classification results of different methods in the Pavia University dataset. (a) Color image; (b) classification result of SVM; (c) classification result of CNN; (d) classification result of our method (Log-Gabor and CNN)
    Fig. 3. Classification results of different methods in the Pavia University dataset. (a) Color image; (b) classification result of SVM; (c) classification result of CNN; (d) classification result of our method (Log-Gabor and CNN)
    Classification results of different method in the Indian Pines dataset. (a) Color image; (b) classification result of SVM; (c) classification result of CNN; (d) classification result of our method (Log-Gabor and CNN)
    Fig. 4. Classification results of different method in the Indian Pines dataset. (a) Color image; (b) classification result of SVM; (c) classification result of CNN; (d) classification result of our method (Log-Gabor and CNN)
    No.ClassNumber of sample
    TrainingTest
    1Asphalt656566
    2Meadows18518464
    3Gravel202079
    4Trees303034
    5Metal sheets151330
    6Bare soil504979
    7Bitumen151315
    8Bricks353647
    9Shadows10937
    Table 1. Number of training and test samples quantity in the Pavia University dataset
    No.ClassNumber of sampleNo.ClassNumber of sample
    TrainingTestTrainingTest
    1Alfalfa10369Oats515
    2Corn-notill140128810Soybean-notill95973
    3Corn-mintill8075011Soybean-mintill2452210
    4Corn2521212Soybean-clean60533
    5Grass-pasture5043213Wheat20185
    6Grass-trees7066014Woods1251140
    7Grass-pasture-mowed101515Building-grass-trees-drives40346
    8Hay-windrowed5042816Stone-steel-towers1578
    Table 2. Number of training and test samples quantity in the Indian Pines dataset
    MethodOA /%AA /%Kappa
    SVM85.9689.570.8940
    CNN96.3296.910.9606
    Log-Gabor and CNN98.4198.050.9833
    Table 3. Classification accuracy of different methods in Pavia University dataset
    MethodOA /%AA /%Kappa
    SVM76.2673.290.7363
    CNN93.9180.120.9217
    Log-Gabor and CNN95.2982.650.9376
    Table 4. Classification accuracy of different methods in Indian Pines dataset
    Qing Fu, Chen Guo, Wenlang Luo. A Hyperspectral Image Classification Method Based on Spectral-Spatial Features[J]. Laser & Optoelectronics Progress, 2020, 57(20): 202803
    Download Citation