• Laser & Optoelectronics Progress
  • Vol. 57, Issue 24, 241016 (2020)
Haoran Hu, Hui Liu*, and Huan Huang
Author Affiliations
  • Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming, Yunnan 650000, China
  • show less
    DOI: 10.3788/LOP57.241016 Cite this Article Set citation alerts
    Haoran Hu, Hui Liu, Huan Huang. Blind Restoration of Multi-Channel Images Based on Total Variation and Dark Pixels[J]. Laser & Optoelectronics Progress, 2020, 57(24): 241016 Copy Citation Text show less
    Comparison of dark pixel sparseness between blurred image and clear image. (a) Clear image; (b) blurred image; (c) dark channel image of Fig.1(a); (d) dark channel image of Fig.1(b)
    Fig. 1. Comparison of dark pixel sparseness between blurred image and clear image. (a) Clear image; (b) blurred image; (c) dark channel image of Fig.1(a); (d) dark channel image of Fig.1(b)
    Target image estimation process
    Fig. 2. Target image estimation process
    Flowchart of PSF estimation algorithm
    Fig. 3. Flowchart of PSF estimation algorithm
    Flowchart of MCBD algorithm
    Fig. 4. Flowchart of MCBD algorithm
    “church” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 5. “church” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    “clock” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 6. “clock” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    “architectural” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 7. “architectural” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    “bridge” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 8. “bridge” blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Real blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 9. Real blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Real blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    Fig. 10. Real blurred images and restoration results. (a1)--(a3) Blurred images; (b) algorithm in Ref.[23]; (c) algorithm in Ref.[6]; (d) proposed algorithm
    PictureAlgorithm in Ref.[23]Algorithm in Ref.[6]Proposed algorithm
    SSIMPSNR /dBSSIMPSNR /dBSSIMPSNR /dB
    “church”0.749226.37440.628424.40550.698825.6721
    “clock”0.628620.80010.553520.12900.651722.0581
    Table 1. Comparison of evaluation indexes of restored images
    PictureAlgorithm in Ref.[23]Algorithm in Ref.[6]Proposed algorithm
    SSIMPSNR /dBSSIMPSNR /dBSSIMPSNR /dB
    “architectural”0.615825.68430.559225.19190.570631.0540
    “bridge”0.771827.85740.607328.17600.609028.3548
    Table 2. Comparison of evaluation indexes of restored images
    Haoran Hu, Hui Liu, Huan Huang. Blind Restoration of Multi-Channel Images Based on Total Variation and Dark Pixels[J]. Laser & Optoelectronics Progress, 2020, 57(24): 241016
    Download Citation