Optics and Precision Engineering, Volume. 32, Issue 16, 2550(2024)

Local region image segmentation by fusion of statistical norm metrics

Gengsheng LI1,2 and Guojun LIU1、*
Author Affiliations
  • 1School of Mathematics and Statistics, Ningxia University, Yinchuan75002, China
  • 2School of Mathematics and Information Technology, Longnan Normal University, Longnan74500, China
  • show less
    Figures & Tables(12)
    Single-phase image segmentation region relationship map
    Framework diagram of the SAB model
    Segmentation results of vascular images
    Segmentation results of vascular images
    Results of lung image segmentation for COVID-19
    Binary segmentation of COVID lung images
    Results of lung image segmentation for COVID-19
    Segmentation results for natural images
    Segmentation results of real images
    • Table 1. Iterations and running time (s) of the images in Fig. 7

      View table
      View in Article

      Table 1. Iterations and running time (s) of the images in Fig. 7

      IndexImage1Image2Image3Image4Image5Image6
      MethodIter/TimesIter/TimesIter/TimesIter/TimesIter/TimesIter/Times
      LBF432/46.203 1345/85.906 3418/67.781 3218/44.859 4332/79.593 8373/38.531 3
      LIF700/3.834 1843/9.326 5792/5.960 2857/9.385 5821/9.016 4800/4.474 8
      FACM140/7.007 1186/13.302 1197/10.721 7217/15.367 2227/15.875 2212/9.876 0
      LSACM60/38.848 886/132.267 866/70.073 376/116.128 660/90.687 168/48.262 5
      ABC225/9.536 9320/30.591 8430/31.581 3420/40.912 7400/39.328 9350/17.102 2
      GFLIF100/6.087 2125/8.295 8130/8.676 7140/9.652 4145/9.768 2150/9.899 0
      HLFRA140/6.552 8175/9.547 6192/9.382 2183/9.982 2200/11.111 9212/9.862 4
      LSE-ADMM400/3.557 3468/7.596 5512/5.655 0475/7.807 2543/8.912 2540/4.160 2
      LPF190/4.397 0410/14.121 3425/8.581 3476/13.620 5389/13.315 9428/7.299 6
      FRAGL240/10.158 2265/13.241 4278/11.931 3252/12.221 2315/15.286 6300/12.357 0
      SAB40/3.137 440/6.639 732/3.786 620/3.035 040/5.822 420/1.831 7
    • Table 2. Comparative results of objective evaluation indicators

      View table
      View in Article

      Table 2. Comparative results of objective evaluation indicators

      Evaluation indicatorsDCFPJCSPTP
      Image IDABCSABABCSABABCSABABCSABABCSAB
      Image10.882 20.980 50.066 80.002 20.789 20.961 70.926 50.997 70.841 90.963 9
      Image20.894 10.985 20.194 90.006 90.808 40.970 80.832 10.993 00.966 00.977 4
      Image30.719 60.987 60.369 80.004 10.562 00.975 60.675 50.995 80.769 90.979 6
      Image40.916 00.986 30.024 40.009 50.844 90.973 10.972 60.990 40.865 60.982 4
      Image50.502 80.985 70.401 30.003 40.335 80.971 70.539 70.996 50.470 50.975 1
    • Table 3. DSC and IoUs of fours images in Fig. 9 by eleven models(DSC/JCS)

      View table
      View in Article

      Table 3. DSC and IoUs of fours images in Fig. 9 by eleven models(DSC/JCS)

      Methods#1#2#3#4
      LBF0.535 10.365 30.533 30.363 60.782 30.643 60.600 90.429 5
      LIF0.157 10.085 20.036 70.018 70.202 50.112 70.210 20.117 4
      FACM0.967 10.936 40.658 20.490 60.921 40.854 20.969 60.941 0
      LSACM0.438 10.280 50.816 10.689 40.877 80.782 20.679 50.514 5
      ABC0.769 00.624 60.900 30.818 70.870 10.770 10.841 60.726 5
      GFLIF0.957 10.917 70.977 40.955 80.859 90.754 20.862 40.758 0
      HLFRA0.947 90.901 00.987 50.975 40.877 80.782 20.878 70.783 6
      LSE-ADMM0.596 90.425 40.958 00.919 40.888 50.799 40.896 80.812 9
      LPF0.152 80.082 70.451 20.291 30.798 90.665 20.660 10.492 7
      FRAGL0.226 20.127 50.822 70.698 90.409 60.257 60.595 60.424 1
      SAB0.990 90.981 90.990 00.980 10.995 00.990 10.995 80.991 7
    Tools

    Get Citation

    Copy Citation Text

    Gengsheng LI, Guojun LIU. Local region image segmentation by fusion of statistical norm metrics[J]. Optics and Precision Engineering, 2024, 32(16): 2550

    Download Citation

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

    Category:

    Received: Mar. 22, 2024

    Accepted: --

    Published Online: Nov. 18, 2024

    The Author Email: LIU Guojun (liugj@nux.edu.cn)

    DOI:10.37188/OPE.20243216.2550

    Topics