Laser & Optoelectronics Progress, Volume. 62, Issue 2, 0237002(2025)

Multiplexed Fusion Deep Aggregate Learning for Underwater Image Enhancement

Yan Chen1,*... Ao Xiao1, Yun Li2, Xiaochun Hu2 and Peiguang Jing3 |Show fewer author(s)
Author Affiliations
  • 1School of Computer, Electronics and Information, Guangxi University, Nanning 530004, Guangxi , China
  • 2School of Big Data and Artificial Intelligence, Guangxi University of Finance and Economics, Nanning 530003, Guangxi , China
  • 3School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, China
  • show less
    Figures & Tables(12)
    Structure of underwater image enhancement model with multiplexed fusion deep aggregate learning
    Loss function curves of proposed method in different training sets
    Treatment results of UIEB dataset images by different methods
    Treatment results of LSUI dataset images by different methods
    Comparison of image sharpness and edges by different methods on LSUI dataset
    Comparison of image details by different methods on UIEB dataset
    Effectiveness comparison of image segmentation and key point detection on UIEB dataset. (a1)‒(c1) Original images; (a2)‒(c2) segmentation images of original images; (a3)‒(c3) key point detection results of original images; (a4)‒(c4) enhanced images; (a5)‒(c5) segmentation images of enhanced images; (a6)‒(c6) key point detection results of enhanced images
    • Table 1. Division of different underwater datasets

      View table

      Table 1. Division of different underwater datasets

      DatasetTotal number of imageTraining setTest set
      UIEB890790100
      LSUI42793879400
    • Table 2. PSNR and SSIM on UIEB and LSUI datasets by different methods

      View table

      Table 2. PSNR and SSIM on UIEB and LSUI datasets by different methods

      MethodUIEBLSUI
      PSNR /dBSSIMPSNR /dBSSIM
      UDCP13.050.6212.680.63
      Fusion17.600.7714.580.76
      Water-Net19.110.7917.820.83
      UGAN20.170.7220.110.78
      Fuine-GAN20.680.8020.640.83
      Ucolor20.630.8421.660.84
      USUIR20.310.8621.920.85
      Proposed22.400.9023.010.86
    • Table 3. UCIQE and entropy on UIEB and LSUI datasets by different methods

      View table

      Table 3. UCIQE and entropy on UIEB and LSUI datasets by different methods

      MethodUIEBLSUI
      UCIQEEntropyUCIQEEntropy
      Fuine-GAN0.8113.820.8913.60
      Ucolor0.6714.180.6813.58
      USUIR0.8214.260.9113.48
      Proposed0.8314.330.9313.62
    • Table 4. Results of ablation experiments

      View table

      Table 4. Results of ablation experiments

      ModulePSNR /dBSSIM
      A20.870.82
      B22.830.85
      C14.610.44
      Proposed23.010.86
    • Table 5. Runtime and FPS of different methods

      View table

      Table 5. Runtime and FPS of different methods

      MethodRuntime /sFPS /(frame·s-1
      Fuine-GAN4.2923.31
      Ucolor45.882.18
      USUIR47.412.11
      Proposed29.133.43
    Tools

    Get Citation

    Copy Citation Text

    Yan Chen, Ao Xiao, Yun Li, Xiaochun Hu, Peiguang Jing. Multiplexed Fusion Deep Aggregate Learning for Underwater Image Enhancement[J]. Laser & Optoelectronics Progress, 2025, 62(2): 0237002

    Download Citation

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

    Category: Digital Image Processing

    Received: Apr. 7, 2024

    Accepted: Apr. 30, 2024

    Published Online: Dec. 17, 2024

    The Author Email:

    DOI:10.3788/LOP241036

    CSTR:32186.14.LOP241036

    Topics