Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Laser & Optoelectronics Progress, Volume. 60, Issue 14, 1410009(2023)

Low-Light Image-Enhancement Algorithm Based on Homomorphic Frequency Division Aggregation

Yali Zhang1, Shaobo Ding1, Changlu Li1、*, Xinming Yao2, and Wenyuan Li1
Author Affiliations
  • 1School of Microelectronics, Tianjin University, Tianjin 300072, China
  • 2Forty-Fifth Research Institute of China Electronics Technology Group Corporation, Beijing 100176, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (15)
    Equations (12)
    References (29)
    Cited By (1)
    Tools
    Paper Information
    Topics
    Figures & Tables(15)
    Network structure diagram of basic residual module

    Fig. 1. Network structure diagram of basic residual module

    Download full size
    Low-light image enhancement algorithm based on homomorphic frequency division aggregation

    Fig. 2. Low-light image enhancement algorithm based on homomorphic frequency division aggregation

    Download full size
    3D diagrams of filter's transfer function. (a) 3D diagram of high-pass filter's transfer function; (b) 3D diagram of low-pass filter's transfer function

    Fig. 3. 3D diagrams of filter's transfer function. (a) 3D diagram of high-pass filter's transfer function; (b) 3D diagram of low-pass filter's transfer function

    Download full size
    Detail enhanced network

    Fig. 4. Detail enhanced network

    Download full size
    Enhancement effects of zebra image using different algorithms

    Fig. 5. Enhancement effects of zebra image using different algorithms

    Download full size
    Enhancement effects of building image using different algorithms

    Fig. 6. Enhancement effects of building image using different algorithms

    Download full size
    Comparison of objective indicators of different algorithms on synthetic dataset

    Fig. 7. Comparison of objective indicators of different algorithms on synthetic dataset

    Download full size
    Enhancement effects of doll image using different algorithms

    Fig. 8. Enhancement effects of doll image using different algorithms

    Download full size
    Enhancement effects of room image using different algorithms

    Fig. 9. Enhancement effects of room image using different algorithms

    Download full size
    Comparison of objective indicators of different algorithms on real datasets

    Fig. 10. Comparison of objective indicators of different algorithms on real datasets

    Download full size
    Output image comparison of missing different modules. (a) Input image; (b) lack of detail enhancement module; (c) lack of brightness enhancement module; (d) lack of local adaptive network; (e) output image of HFNet; (f) reference image

    Fig. 11. Output image comparison of missing different modules. (a) Input image; (b) lack of detail enhancement module; (c) lack of brightness enhancement module; (d) lack of local adaptive network; (e) output image of HFNet; (f) reference image

    Download full size

    • Table 1. Quantitative comparison of synthetic low-light images obtained by different algorithms

      View table

      Table 1. Quantitative comparison of synthetic low-light images obtained by different algorithms

      AlgorithmLIMELLNetRetinexNetZero-DCEEnlightenGANHFNet
      PSNR24.9526.0928.3628.6228.9728.84
      SSIM0.6200.7180.7890.8240.8510.862

    • Table 2. Comparison of subjective evaluation on synthetic low-light images

      View table

      Table 2. Comparison of subjective evaluation on synthetic low-light images

      Image numberLIMELLNetRetinexNetZero-DCEEnlightenGANHFNet
      12.952.703.053.153.203.35
      23.002.852.903.103.253.30
      32.952.652.853.053.203.10
      42.502.602.753.002.953.05
      52.952.902.853.053.003.15
      62.902.953.053.103.153.25
      72.852.702.803.002.953.10
      83.002.802.853.053.103.25

    • Table 3. Quantitative comparison of different algorithms on real low-light images

      View table

      Table 3. Quantitative comparison of different algorithms on real low-light images

      AlgorithmLIMELLNetRetinexNetZero-DCEEnlightenGANHFNet
      IE7.20126.44246.73286.89506.83367.0159
      NIQE6.98017.62437.44056.85726.64916.7696
      NIQMC5.01224.16914.38194.89274.66075.1008

    • Table 4. Comparison of subjective evaluation on real low-light images

      View table

      Table 4. Comparison of subjective evaluation on real low-light images

      Image numberLIMELLNetRetinexNetZero-DCEEnlightenGANHFNet
      12.902.753.003.103.153.25
      23.052.802.953.103.253.30
      32.902.752.953.153.203.25
      42.952.802.853.003.053.20
      53.052.952.803.103.253.15
      63.002.903.053.153.253.30
      72.952.802.753.203.153.25
      83.002.852.903.153.103.20
    Tools

    Get Citation

    Copy Citation Text

    Yali Zhang, Shaobo Ding, Changlu Li, Xinming Yao, Wenyuan Li. Low-Light Image-Enhancement Algorithm Based on Homomorphic Frequency Division Aggregation[J]. Laser & Optoelectronics Progress, 2023, 60(14): 1410009

    Download Citation

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

    Category: Image Processing

    Received: Mar. 22, 2022

    Accepted: Sep. 5, 2022

    Published Online: Jul. 17, 2023

    The Author Email: Li Changlu (3383928167@qq.com)

    DOI:10.3788/LOP221078

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology