Journal of Applied Optics, Volume. 45, Issue 4, 819(2024)

Low-level-light image enhancement method based on Retinex theory by improving dual components

Huateng CHEN1, Lei LIU1、*, Yunsheng QIAN1, Weitao DENG1, and Feng SHI2
Author Affiliations
  • 1School of Electronic Engineering and Optoelectronic Technology, Nanjing University of Science and Technology, Nanjing 210094, China
  • 2Key Laboratory of Low-Light-Level Night Vision Technology, Xi'an 710065, China
  • show less
    Figures & Tables(14)
    Example of Retinex model method based on two component enhancement
    Brightness enhancement at different a values
    Illumination component enhancement results
    Effect of multi-scale detail enhancement of reflection component
    Contrast enhancement processing
    Comparison of different low-level-light enhancement methods
    Comparison of enhancement results of 35 images dataset images
    Comparison of enhancement results of 200 Darkface dataset images
    Comparison of enhancement results of LOL dataset images
    Enhancement effect of low-level-light images captured by cameras
    • Table 1. NIQE comparison between Figure 7~Figure 9

      View table
      View in Article

      Table 1. NIQE comparison between Figure 7~Figure 9

      图像HECALHESSRMSRCRDongLIMECRM本文方法
      图73.7953.3534.4013.6153.2293.1683.4633.341
      图82.7122.8093.8873.5662.3483.0502.6862.151
      图93.7423.6324.2813.7383.5133.8453.5633.425
    • Table 2. Average NIQE of 35 images and 200 Darkface dataset images

      View table
      View in Article

      Table 2. Average NIQE of 35 images and 200 Darkface dataset images

      数据集HECALHESSRMSRCRDongLIMECRM本文方法
      35 images3.8593.6163.8513.4533.2373.2543.3513.141
      200 Darkface2.7032.6573.7953.4352.7842.9092.6312.362
    • Table 3. Quantitative comparison of LOL dataset in PSNR and SSIM

      View table
      View in Article

      Table 3. Quantitative comparison of LOL dataset in PSNR and SSIM

      指标HECALHESSRMSRCRDongLIMECRM本文方法
      PSNR12.30114.24615.64115.72516.31716.59916.57417.741
      SSIM0.4480.5940.6620.6820.6970.6640.6590.765
    • Table 4. Comparison of calculation time for different low-level-light image enhancement methods s

      View table
      View in Article

      Table 4. Comparison of calculation time for different low-level-light image enhancement methods s

      指标HECALHESSRMSRCRDongLIMECRM本文方法
      PNSR1.212.781.482.071.564.733.412.56
    Tools

    Get Citation

    Copy Citation Text

    Huateng CHEN, Lei LIU, Yunsheng QIAN, Weitao DENG, Feng SHI. Low-level-light image enhancement method based on Retinex theory by improving dual components[J]. Journal of Applied Optics, 2024, 45(4): 819

    Download Citation

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

    Category: Research Articles

    Received: Jun. 21, 2023

    Accepted: --

    Published Online: Oct. 21, 2024

    The Author Email: Lei LIU (刘磊)

    DOI:10.5768/JAO202445.0404001

    Topics