• Infrared and Laser Engineering
  • Vol. 51, Issue 9, 20210868 (2022)
Anqi Li1, Li Ma1、2、*, Helong Yu1、2、*, and Hanbo Zhang1
Author Affiliations
  • 1College of Information Technology, Jilin Agricultural University, Changchun 130118, China
  • 2Smart Agriculture Research Institute, Jilin Agricultural University, Changchun 130118, China
  • show less
    DOI: 10.3788/IRLA20210868 Cite this Article
    Anqi Li, Li Ma, Helong Yu, Hanbo Zhang. Research on the classification of typical crops in remote sensing images by improved U-Net algorithm[J]. Infrared and Laser Engineering, 2022, 51(9): 20210868 Copy Citation Text show less
    Location diagram of the study area
    Fig. 1. Location diagram of the study area
    Part of the label data set
    Fig. 2. Part of the label data set
    Structure of ASPP
    Fig. 3. Structure of ASPP
    Classification structure of crops based on improved U-Net
    Fig. 4. Classification structure of crops based on improved U-Net
    Comparison of the experimental results of proposed method and other algorithms
    Fig. 5. Comparison of the experimental results of proposed method and other algorithms
    Deep learning segmentation modelModel characteristics
    FCNFor the first time, a fully convolutional network based on the end-to-end concept is proposed, which removes the fully connected layer and samples in the deconvolutional layer
    SegNetThe pooling layer result is used in the decoding and a large amount of coding information is introduced
    U-NetBased on the end-to-end standard network structure, the decoder is obtained by splicing the results of each layer on the encoder, and the result is more ideal
    Improve U-NetThe ability of semantic recognition is enhanced, and it is more sensitive to feature extraction
    Table 1. Characteristics of network model for different deep learning
    Experimental network U-NetSegNetFCNImprove U-Net
    OAMIoUOAMIoUOAMIoUOAMIoU
    Precision85.41%0.3984.86%0.3986.44%0.4588.33%0.52
    Table 2. Experimental results of crop recognition with different methods
    Anqi Li, Li Ma, Helong Yu, Hanbo Zhang. Research on the classification of typical crops in remote sensing images by improved U-Net algorithm[J]. Infrared and Laser Engineering, 2022, 51(9): 20210868
    Download Citation