Optics and Precision Engineering, Volume. 30, Issue 4, 464(2022)

Single image dehazing with sky segmentation and haze density estimation

Jianwei LV1...2, Feng QIAN1, Haonan HAN1,2 and Bao ZHANG1,* |Show fewer author(s)
Author Affiliations
  • 1Changchun Institute of Optics,Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun30033, China
  • 2University of Chinese Academy of Sciences, Beijing100049, China
  • show less
    Figures & Tables(12)
    Block diagram of algorithm principle
    Results of sky segmentation.
    Hazy image and its atmosphere light estimation results
    Parts of haze-free images and the corresponding synthetic hazy images
    Results of haze density estimation and image dehazing
    Results of hazy images transmission and dehazing
    Comparison of the dehazing methods without sky
    Comparison of the dehazing methods with sky
    Hazy images and the corresponding ground truth images
    • Table 1. Comparison of the parameters of image quality assessment

      View table
      View in Article

      Table 1. Comparison of the parameters of image quality assessment

      有雾图像算法erH
      1Zhu算法1.583 51.028 10.02
      He算法3.077 81.639 20.02
      Beman算法5.567 11.562 10.04
      Cai算法1.240 81.167 20.02
      本文算法4.131 71.363 10.04
      2Zhu算法1.865 40.737 50.01
      He算法4.074 21.281 40.06
      Beman算法4.121 91.562 10.08
      Cai算法1.005 81.049 30.04
      本文算法4.823 31.681 30.01
      3Zhu算法1.313 41.617 60.01
      He算法1.509 31.960 80.01
      Beman算法5.192 82.086 30.05
      Cai算法4.863 91.448 20.02
      本文算法6.328 54.811 70.01
      4Zhu算法4.009 91.139 60.02
      He算法4.463 31.120 60.02
      Beman算法5.488 31.588 20.04
      Cai算法1.030 91.079 20.01
      本文算法6.821 61.651 70.05
    • Table 2. Comparison of PSNR and SSIM values

      View table
      View in Article

      Table 2. Comparison of PSNR and SSIM values

      有雾图像算法PSNRSSIM
      1Zhu算法7.140 20.506 9
      He算法6.318 90.504 4
      Beman算法8.711 70.521 9
      Cai算法7.638 10.484 9
      本文算法8.638 20.518 2
      2Zhu算法9.950 30.677 2
      He算法11.632 90.712 5
      Beman算法13.948 80.771 2
      Cai算法9.749 60.661 1
      本文算法15.142 30.754 5
      3Zhu算法14.941 40.706 5
      He算法15.968 60.723 8
      Beman算法15.455 90.701 7
      Cai算法14.674 50.698 6
      本文算法16.559 70.727 1
      4Zhu算法11.470 10.232 3
      He算法12.085 80.217 2
      Beman算法10.681 60.174 3
      Cai算法11.343 20.227 4
      本文算法12.187 80.235 4
    • Table 3. Comparison of running time of image dehazing algorithms

      View table
      View in Article

      Table 3. Comparison of running time of image dehazing algorithms

      图像大小算法运行时间
      1 024×768Zhu算法5.89s
      He算法10.15s
      Beman算法19.46s
      Cai算法20.37s
      本文算法8.05s
      600×400Zhu算法2.17s
      He算法3.71s
      Beman算法5.49s
      Cai算法6.63s
      本文算法3.42s
      474×406Zhu算法2.26s
      He算法3.01s
      Beman算法3.25s
      Cai算法4.34s
      本文算法2.72s
      580×435Zhu算法1.93s
      He算法3.39s
      Beman算法6.02s
      Cai算法6.19s
      本文算法2.41s
    Tools

    Get Citation

    Copy Citation Text

    Jianwei LV, Feng QIAN, Haonan HAN, Bao ZHANG. Single image dehazing with sky segmentation and haze density estimation[J]. Optics and Precision Engineering, 2022, 30(4): 464

    Download Citation

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

    Category: Information Sciences

    Received: Jun. 3, 2021

    Accepted: --

    Published Online: Mar. 4, 2022

    The Author Email: ZHANG Bao (zhangb@ciomp.edu.cn)

    DOI:10.37188/OPE.20223004.0464

    Topics