Optics and Precision Engineering, Volume. 32, Issue 10, 1567(2024)

Infrared image and visible image fusion algorithm based on secondary image decomposition

Xin MA... Chunyu YU*, Yixin TONG and Jun ZHANG |Show fewer author(s)
Author Affiliations
  • College of Electronic and Optical Engineering & College of Flexible Electronics (Future Technology),Nanjing University of Posts and Telecommunications,Nanjing210023,China
  • show less
    Figures & Tables(16)
    Training network frame diagram
    Test network frame diagram
    BERM module
    DERM module
    Dual element-wise attention mechanism module
    Channel attention diagram
    Various methods fuse images on TNO datasets
    TNO data set objective evaluation indicators line chart
    Analysis of RoadScene data set fusion results
    RoadScene data set objective evaluation index line chart
    Ablation results
    • Table 1. Comparison of fusion results of three fusion strategies in TNO dataset

      View table
      View in Article

      Table 1. Comparison of fusion results of three fusion strategies in TNO dataset

      ENSDSFVIFAGSCD
      像素相加法7.338 99.843 00.046 80.828 04.529 41.927 8
      加权相加法6.742 49.116 20.027 70.698 02.725 81.695 2
      通道注意力7.337 19.816 40.047 00.812 64.445 71.864 4
    • Table 2. Comparison of fusion results of three fusion strategies in RoadScene dataset

      View table
      View in Article

      Table 2. Comparison of fusion results of three fusion strategies in RoadScene dataset

      ENSDSFVIFAGSCD
      像素相加法7.377 810.929 60.055 60.715 45.452 31.809 2
      加权相加法6.920 39.715 20.035 60.651 93.615 51.459 2
      通道注意力7.357 311.247 60.056 00.690 75.303 41.796 3
    • Table 3. Ten fusion methods were objectively compared on the TNO dataset

      View table
      View in Article

      Table 3. Ten fusion methods were objectively compared on the TNO dataset

      ENSDSFVIFAGSCDMS_SSIMQabf
      DenseFuse6.538 59.028 20.024 60.667 92.446 71.633 60.872 40.331 3
      RFN-Nest7.081 49.637 60.021 70.792 42.495 81.817 60.906 10.327 9
      IFCNN6.889 79.342 00.046 60.784 74.497 81.699 30.909 30.504 2
      NestFuse7.037 29.521 70.036 30.863 33.482 21.729 80.880 10.531 7
      GTF6.975 29.652 40.032 90.606 83.116 40.980 90.809 80.319 4
      Dualbranch6.523 48.989 30.023 00.661 92.377 41.624 40.867 60.321 2
      LRRNet7.160 09.317 30.035 60.785 93.617 01.618 80.866 40.358 6
      SMOA6.724 69.193 30.025 10.720 82.461 71.715 00.885 70.306 1
      SDNet6.722 89.505 80.035 60.966 23.393 60.880 70.726 70.428 9
      SIDFuse7.338 99.843 00.046 80.828 04.529 41.907 80.931 20.432 9
    • Table 4. Ten fusion methods were objectively compared on RoadScene dataset

      View table
      View in Article

      Table 4. Ten fusion methods were objectively compared on RoadScene dataset

      ENSDSFVIFAGSCDMS_SSIMQabf
      GTF7.394 810.474 50.037 30.573 63.120 91.009 70.732 30.337 6
      DenseFuse6.802 59.403 30.036 90.638 83.708 21.391 50.848 10.353 1
      RFN-Nest7.299 29.958 80.030 70.691 23.389 41.727 20.863 90.292 6
      IFCNN6.911 69.470 30.054 90.639 35.228 11.439 00.900 10.521 8
      NestFuse7.396 510.329 60.053 10.824 45.075 31.707 10.865 00.502 2
      Dualbranch6.784 69.388 70.032 70.640 13.362 61.386 20.856 50.418 0
      LRRnet7.132 89.926 80.049 00.673 74.718 31.594 70.788 80.348 1
      SMOA6.957 69.641 10.041 20.672 94.093 51.539 20.877 00.432 2
      SDNet7.232 19.813 30.058 30.737 55.994 51.444 80.893 10.495 3
      SIDFuse7.487 610.796 40.060 10.728 25.969 91.837 10.911 70.488 4
    • Table 5. TNO data set ablation experiment

      View table
      View in Article

      Table 5. TNO data set ablation experiment

      ENSDSFVIFAGSCD
      实验a7.004 38.876 30.040 70.766 44.407 41.797 6
      实验b7.032 98.964 10.044 70.776 24.175 11.815 0
      实验c7.220 49.472 50.045 40.816 04.362 01.875 8
      实验d7.256 29.715 20.047 40.821 64.801 91.901 4
      实验e7.338 99.843 00.046 80.828 04.529 41.907 8
    Tools

    Get Citation

    Copy Citation Text

    Xin MA, Chunyu YU, Yixin TONG, Jun ZHANG. Infrared image and visible image fusion algorithm based on secondary image decomposition[J]. Optics and Precision Engineering, 2024, 32(10): 1567

    Download Citation

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

    Category:

    Received: Sep. 26, 2023

    Accepted: --

    Published Online: Jul. 8, 2024

    The Author Email: YU Chunyu (yucy@njupt.edu.cn)

    DOI:10.37188/OPE.20243210.1567

    Topics