Laser & Optoelectronics Progress, Volume. 61, Issue 12, 1237005(2024)

High Dynamic Range Image Reconstruction Based on Dual-Attention Network

Xianfeng Wang1,2,3, Shiben Liu2,3, Jiandong Tian2,3、*, Juanping Zhao1, yajing Liu2,3, and Chunhui Hao1,2,3
Author Affiliations
  • 1College of Information Engineering, Shenyang University of Chemical Technology, Shenyang 110142, Liaoning, China
  • 2State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, Liaoning, China
  • 3Institute of Robotics and Intelligent Manufacturing Innovation, Chinese Academy of Sciences, Shenyang 110169, Liaoning, China
  • show less
    Figures & Tables(6)
    Overall of network structure
    Flowchart of pyramidal dilation convolution
    Comparison of experimental results
    Comparison of ablation experiment effect
    • Table 1. Quantitative evaluation results of experiments

      View table

      Table 1. Quantitative evaluation results of experiments

      MethodPSNRSSIMLPIPS
      DRHDR17.0520.7850.164
      ExpandNet16.2390.5940.200
      U2Fusion17.2960.7430.159
      Method in Ref.[2318.3770.8530.143
      Ours18.4960.8820.149
    • Table 2. Quantitative evaluation results of ablation experiments

      View table

      Table 2. Quantitative evaluation results of ablation experiments

      MethodPSNRSSIMLPIPS
      Without DAM17.5110.7680.257
      Without FEM17.0370.7590.205
      Without Dilated Conv17.2960.7400.201
      Ours18.4960.8820.149
    Tools

    Get Citation

    Copy Citation Text

    Xianfeng Wang, Shiben Liu, Jiandong Tian, Juanping Zhao, yajing Liu, Chunhui Hao. High Dynamic Range Image Reconstruction Based on Dual-Attention Network[J]. Laser & Optoelectronics Progress, 2024, 61(12): 1237005

    Download Citation

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

    Category: Digital Image Processing

    Received: Jul. 21, 2023

    Accepted: Sep. 6, 2023

    Published Online: May. 29, 2024

    The Author Email: Jiandong Tian (tianjd@sia.cn)

    DOI:10.3788/LOP231770

    CSTR:32186.14.LOP231770

    Topics