Infrared and Laser Engineering, Volume. 54, Issue 6, 20240600(2025)

Imagedefogging algorithm based on transmittance multiple guidance and sharpening compensation

Dongyang SHI1, Sheng HUANG1,2, Liu YANG1, and Liujiang GUO1
Author Affiliations
  • 1School of Communications and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
  • 2Key Laboratory of Optical Communications and Network, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
  • show less
    Figures & Tables(16)
    Imagedefogging algorithm based on transmittance multiple guidance and sharpening compensation
    Atmospheric light value solving module based on threshold segmentation
    Dehazing visualization results in mist scene
    Visualization results of the image before sharpening and after adding the sharpening module
    De-fogging visualization results before and after brightness compensation
    De-fogging visualization results for medium haze scenarios
    Visualization results of fog removal in dense fog scene
    De-fogging visualization results for real haze scenarios
    • Table 1. Bright area transmittance and non-bright area transmittance

      View table
      View in Article

      Table 1. Bright area transmittance and non-bright area transmittance

      Different regions of the imageTransmittance and brightness
      Bright area0.18
      Non-bright area0.97
      Mean brightness Bg0.56
    • Table 2. Performance evaluation results of the defogging model under the misty dataset

      View table
      View in Article

      Table 2. Performance evaluation results of the defogging model under the misty dataset

      PerformanceRetinexHEDADCP[8]PSD[20]Dehamer[22]C2Pnet[23]DCMPNet[25]DTGSC
      MSE52.4932.2638.7523.4265.1530.9221.7010.75
      PSNR/dB31.0534.2333.3236.1331.0234.1635.9940.67
      SSIM78.11%69.81%70.96%70.98%70.55%87.96%90.66%91.99%
    • Table 3. Image defogging performance before sharpening and after adding mainstream sharpening module

      View table
      View in Article

      Table 3. Image defogging performance before sharpening and after adding mainstream sharpening module

      PerformanceBefore sharpening+ LSF+ UMF+ SFF+ VRF
      MSE10.7513.0613.7113.765.67
      PSNR/dB40.6738.0437.6537.6341.21
      SSIM91.99%94.43%91.85%91.87%70.16%
    • Table 4. Image defogging performance before and after brightness compensation

      View table
      View in Article

      Table 4. Image defogging performance before and after brightness compensation

      Before and after brightness compensationMSEPSNR/dBSSIM
      Master model10.7540.6791.99%
      After image sharpening13.0638.0494.43%
      After brightness compensation8.0641.1394.02%
    • Table 5. Performance evaluation results of the defogging models under medium haze dataset

      View table
      View in Article

      Table 5. Performance evaluation results of the defogging models under medium haze dataset

      PerformanceRetinexHEDADCP[8]PSD[20]Dehamer[22]C2Pnet[23]DCMPNet[25]DTGSC
      MSE58.4134.0545.1622.5176.0240.9627.4813.06
      PSNR/dB31.4134.2932.3535.7430.1932.7334.8039.70
      SSIM79.75%73.03%74.25%69.54%74.62%85.84%88.54%89.80%
    • Table 6. Performance evaluation results of the defogging models under dense fog dataset

      View table
      View in Article

      Table 6. Performance evaluation results of the defogging models under dense fog dataset

      PerformanceRetinexHEDADCP[8]PSD[20]Dehamer[22]C2Pnet[23]DCMPNet[25]DTGSC
      MSE103.5854.2569.5429.25130.7948.7840.5313.71
      PSNR/dB29.3232.2330.9834.4428.2332.6733.5438.89
      SSIM79.81%72.32%77.87%73.73%72.63%84.68%87.15%86.29%
    • Table 7. Performance evaluation results of the defogging models in real haze scenarios

      View table
      View in Article

      Table 7. Performance evaluation results of the defogging models in real haze scenarios

      PerformanceRetinexHEDADCP[8]PSD[20]Dehamer[22]C2Pnet[23]DCMPNet[25]DTGSC
      MSE299.84177.4843.0359.58337.70102.9036.716.75
      PSNR/dB23.3625.6431.7930.3822.8528.0132.4839.84
      SSIM72.72%63.66%54.23%55.31%70.99%62.16%74.91%83.25%
    • Table 8. Mean running time of different defogging models with misty dataset

      View table
      View in Article

      Table 8. Mean running time of different defogging models with misty dataset

      Defogging modelDefogging time/s
      DCP[8]0.58
      C2Pnet[23]0.87
      DCMPNet[25]0.71
      DTGSC0.63
    Tools

    Get Citation

    Copy Citation Text

    Dongyang SHI, Sheng HUANG, Liu YANG, Liujiang GUO. Imagedefogging algorithm based on transmittance multiple guidance and sharpening compensation[J]. Infrared and Laser Engineering, 2025, 54(6): 20240600

    Download Citation

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

    Category: Optical imaging, display and information processing

    Received: Dec. 23, 2024

    Accepted: --

    Published Online: Jul. 1, 2025

    The Author Email:

    DOI:10.3788/IRLA20240600

    Topics