• Laser & Optoelectronics Progress
  • Vol. 57, Issue 17, 172802 (2020)
Qing Fu1、2、3, Wenlang Luo1、2、*, and Jingxiang Lü1、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/LOP57.172802 Cite this Article Set citation alerts
    Qing Fu, Wenlang Luo, Jingxiang Lü. Land Utilization Change Detection of Satellite Remote Sensing Image Based on AlexNet and Support Vector Machine[J]. Laser & Optoelectronics Progress, 2020, 57(17): 172802 Copy Citation Text show less
    AlexNet architecture
    Fig. 1. AlexNet architecture
    Flow chart of land use classification based on AlexNet and SVM
    Fig. 2. Flow chart of land use classification based on AlexNet and SVM
    Distribution of validation samples
    Fig. 3. Distribution of validation samples
    Classification accuracy of the model at different window sizes
    Fig. 4. Classification accuracy of the model at different window sizes
    Classification results of different algorithms. (a) Maximum likelihood algorithm; (b) SVM algorithm; (c) AlexNet algorithm; (d) our algorithm (AlexNet+SVM)
    Fig. 5. Classification results of different algorithms. (a) Maximum likelihood algorithm; (b) SVM algorithm; (c) AlexNet algorithm; (d) our algorithm (AlexNet+SVM)
    Classification accuracy of different algorithms
    Fig. 6. Classification accuracy of different algorithms
    Dynamic change of land use types in Nanchang. (a) From 2013 to 2014; (b) from 2013 to 2015; (c) from 2013 to 2016; (d) from 2013 to 2017
    Fig. 7. Dynamic change of land use types in Nanchang. (a) From 2013 to 2014; (b) from 2013 to 2015; (c) from 2013 to 2016; (d) from 2013 to 2017
    Bandrange /μmSpatialresolution /mWidth /kmRevisittime /d
    0.45-0.52
    0.52-0.59168002
    0.63-0.69
    0.77-0.89
    Table 1. GF-1 multispectral camera technical parameters
    No.NameDetail
    1VegetationIncluding forest land, grassland and other green vegetation covered land
    2BuildingIncluding urban land, residential land and traffic land
    3WaterIncluding rivers, lakes, reservoirs, ponds and ditches
    4Bare landIncluding natural bare land, developing bare land and beach sand
    Table 2. Land use/cover change classification system
    Sample size /(pixel×pixel)Sample data numbers
    VegetationWaterBare landBuilding
    5×5900800700650
    7×7900800700650
    9×9900800700650
    11×11900800700650
    13×13900800700650
    Table 3. Training sample data
    20132017
    BuildingBare landVegetationWater
    Building149.457.569.458.78
    Bare land7.58268.2110.345.66
    Vegetation22.699.983489.1239.32
    Water10.458.5937.41634.74
    Table 4. Land use transition matrix in Nanchang from 2013 to 2017km2
    CategoryLand use area change /km2
    From 2013to 2014From 2013to 2015From 2013to 2016From 2013to 2017
    Vegetation-44.2-38.40-45.03-54.74
    Water28.6419.5610.1222.12
    Building11.4614.5217.8319.45
    Bare land20.18.327.065.17
    Table 5. Statistics of land use area change in Nanchang from 2013 to 2017
    Qing Fu, Wenlang Luo, Jingxiang Lü. Land Utilization Change Detection of Satellite Remote Sensing Image Based on AlexNet and Support Vector Machine[J]. Laser & Optoelectronics Progress, 2020, 57(17): 172802
    Download Citation