• Laser & Optoelectronics Progress
  • Vol. 57, Issue 6, 061011 (2020)
Qiang Gao, Liaolin Hu*, and Xin Chen
Author Affiliations
  • School of Mechanical and Precision Instrument Engineering, Xi'an University of Technology, Xi'an, Shaanxi 710048, China
  • show less
    DOI: 10.3788/LOP57.061011 Cite this Article Set citation alerts
    Qiang Gao, Liaolin Hu, Xin Chen. Image Dehazing Method Based on Dark Channel Compensation and Improvement of Atmospheric Light Value[J]. Laser & Optoelectronics Progress, 2020, 57(6): 061011 Copy Citation Text show less
    Comparison of dark channel. (a) Haze image; (b) original dark channel; (c) compensated dark channel; (d) original result; (e) compensated result
    Fig. 1. Comparison of dark channel. (a) Haze image; (b) original dark channel; (c) compensated dark channel; (d) original result; (e) compensated result
    Quadtree partition region. (a) Partitioned area; (b) neglected area
    Fig. 2. Quadtree partition region. (a) Partitioned area; (b) neglected area
    Proposed segmentation method. (a) Partitioned area; (b) combined image
    Fig. 3. Proposed segmentation method. (a) Partitioned area; (b) combined image
    Atmospheric light value selection. (a) Haze image; (b) method of Ref. [4]; (c) proposed method; (d) result of Ref. [4]; (e) result of proposed method
    Fig. 4. Atmospheric light value selection. (a) Haze image; (b) method of Ref. [4]; (c) proposed method; (d) result of Ref. [4]; (e) result of proposed method
    Flow of the proposed method
    Fig. 5. Flow of the proposed method
    Four sets of image processing results. (a) Haze images; (b) method of Ref. [6]; (c) method of Ref. [7]; (d) method of Ref. [8]; (e) proposed method
    Fig. 6. Four sets of image processing results. (a) Haze images; (b) method of Ref. [6]; (c) method of Ref. [7]; (d) method of Ref. [8]; (e) proposed method
    ImageAverage gradient
    Ref. [6]Ref. [7]Ref. [8]Proposed
    Frist set8.747910.48689.68488.8726
    Second set5.28088.45157.44288.4428
    Third set3.26206.27976.71706.7210
    Fourth set5.65178.35036.01478.2042
    Table 1. Average gradient comparison of four image sets
    ImageContrast
    Ref. [6]Ref. [7]Ref. [8]Proposed
    Frist set22.142329.090426.862133.9113
    Second set19.157620.471848.599319.6998
    Third set21.362529.030844.133837.7893
    Fourth set25.421927.452738.032637.9498
    Table 2. Contrast comparison of four image sets
    ImageInformation entropy /bit
    Ref. [6]Ref. [7]Ref. [8]Proposed
    Frist set15.687516.157515.872615.5319
    Second set14.606615.067914.674815.0226
    Third set15.131915.070215.207715.1212
    Fourth set13.015314.523515.234514.6263
    Table 3. Information entropy comparison of four image sets
    ImageRunning time /s
    Ref. [6]Ref. [7]Ref. [8]Proposed
    Frist set0.89782.20940.47241.2991
    Second set0.87502.04790.27101.2015
    Third set1.41112.35380.40971.4659
    Fourth set1.60233.17340.45721.6046
    Table 4. Running time of different algorithms
    Qiang Gao, Liaolin Hu, Xin Chen. Image Dehazing Method Based on Dark Channel Compensation and Improvement of Atmospheric Light Value[J]. Laser & Optoelectronics Progress, 2020, 57(6): 061011
    Download Citation