Acta Photonica Sinica, Volume. 51, Issue 4, 0410001(2022)

Polarization Image Super-resolution Reconstruction Based on Dual Attention Residual Network

Guoming XU1...2,3, Jie WANG1,*, Jian MA1,2, Yong WANG3, Jiaqing LIU1 and Yi LI4 |Show fewer author(s)
Author Affiliations
  • 1School of Internet,Anhui University,Hefei 230039,China
  • 2National Engineering Research Center for Agro-Ecological Big Data Analysis & Application,Anhui University,Hefei 230601,China
  • 3Anhui Province Key Laboratory of Polarized Imaging Detecting Technology,Army Artillery and Air Defense Forces Academy of PLA,Hefei 230031,China
  • 4Institute of Intelligent Technology,Anhui Wenda University of Information Engineering,Hefei 231201,China
  • show less
    Figures & Tables(18)
    Basic structure of residual block
    Network structure of channel attention mechanism
    Spatial attention mechanism network structure
    Network architecture of dual attention deep residual network
    Network structure of base modules
    Network structure of dual attention residual block
    Network structure of spatial attention block
    Spectral polarization camera
    Different polarization direction of fabric image
    Reconstruction effect of truck contraction model by different methods
    Reconstruction effect of airplane image by different methods
    Analysis of monument image full polarization
    Reconstruction effect of image V by different methods
    Comparison between polarization image reconstruction results of camouflage net and the system-acquired results
    Curve of relationship between no reference evaluation index and supercell multiple
    • Table 1. Comparison of indicators of fabric image polarization by different algorithms

      View table
      View in Article

      Table 1. Comparison of indicators of fabric image polarization by different algorithms

      ScaleMethodParams60°120°
      PSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIM
      2Bicubic-37.626 00.939 334.775 40.925 436.116 40.929 1
      SRCNN57K39.808 20.959 037.164 20.950 738.393 50.952 4
      FSRCNN12K39.858 20.959 937.520 40.952 438.583 10.953 8
      EDSR40.7M42.277 40.968 639.892 00.962 240.974 90.963 6
      Ours17.7M42.294 50.968 839.919 20.962 640.988 00.963 9
      3Bicubic-34.056 60.876 131.197 20.846 132.629 70.857 4
      SRCNN57K35.727 50.902 932.984 90.881 034.356 10.888 8
      FSRCNN12K35.572 20.902 733.047 50.881 834.280 40.889 0
      EDSR43.6M37.783 20.923 635.145 50.905 036.403 20.911 5
      Ours17.9M37.834 80.924 635.245 30.906 336.494 20.912 8
      4Bicubic-31.966 80.824 329.106 60.780 030.555 30.798 3
      SRCNN57K33.382 40.848 730.478 20.811 431.910 60.826 3
      FSRCNN12K33.471 30.849 330.665 40.813 732.031 80.827 7
      EDSR43M35.272 30.881 632.495 80.850 533.794 30.862 4
      Ours17.9M35.356 80.882 532.563 50.851 933.863 40.863 5
    • Table 2. Comparison of indicators on the analysis of monument image full polarization by different algorithms

      View table
      View in Article

      Table 2. Comparison of indicators on the analysis of monument image full polarization by different algorithms

      MethodTime/sImage VImage EyImage S0Image QImage U
      PSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIM
      Bicubic0.0522.406 90.907 022.043 20.909 252.669 60.997 520.953 30.783 822.587 60.909 6
      SRCNN0.2427.944 30.963 626.216 70.965 951.989 40.997 623.089 10.870 928.221 50.964 3
      FSRCNN0.2527.567 70.960 425.850 30.958 851.167 90.997 324.194 10.903 827.786 70.961 9
      EDSR0.4932.761 50.990 829.315 70.982 153.533 40.997 629.268 50.958 532.860 40.990 9
      Ours0.4835.331 90.993 930.940 40.987 153.538 10.997 629.141 30.956 135.052 70.993 7
    • Table 3. No reference comparison of camouflage net polarization images results

      View table
      View in Article

      Table 3. No reference comparison of camouflage net polarization images results

      ScaleInformation entropyDefinition
      LRHRSRLRHRSR
      ×26.836.876.850.910.930.93
      6.576.616.590.910.930.94
      6.726.766.740.910.930.94
    Tools

    Get Citation

    Copy Citation Text

    Guoming XU, Jie WANG, Jian MA, Yong WANG, Jiaqing LIU, Yi LI. Polarization Image Super-resolution Reconstruction Based on Dual Attention Residual Network[J]. Acta Photonica Sinica, 2022, 51(4): 0410001

    Download Citation

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

    Category:

    Received: Sep. 18, 2021

    Accepted: Nov. 17, 2021

    Published Online: May. 18, 2022

    The Author Email: WANG Jie (1197545193@qq.com)

    DOI:10.3788/gzxb20225104.0410001

    Topics