• Acta Optica Sinica
  • Vol. 38, Issue 11, 1110003 (2018)
Chenggang Dai1、2、3、*, Mingxing Lin1、2、3、*, Zhen Wang1、2、3, Dong Zhang4, and Zhiguang Guan5
Author Affiliations
  • 1 School of Mechanical Engineering, Shandong University, Jinan, Shandong 250061, China
  • 2 National Demonstration Center for Experimental Mechanical Engineering, School of Mechanical Engineering, Shandong University, Jinan, Shandong 250061, China
  • 3 Key Laboratory of High-efficiency and Clean Mechanical Manufacture of Ministry of Education, School of Mechanical Engineering, Shandong University, Jinan, Shandong 250061, China
  • 4 Institute of Automation Shangdong Academy of Sciences, Jinan, Shandong 250013, China
  • 5 School of Construction Machinery, Shandong Jiaotong University, Jinan, Shandong 250023, China
  • show less
    DOI: 10.3788/AOS201838.1110003 Cite this Article Set citation alerts
    Chenggang Dai, Mingxing Lin, Zhen Wang, Dong Zhang, Zhiguang Guan. Color Compensation Based on Bright Channel and Fusion for Underwater Image Enhancement[J]. Acta Optica Sinica, 2018, 38(11): 1110003 Copy Citation Text show less
    Flow chart of the proposed method
    Fig. 1. Flow chart of the proposed method
    Schematic of underwater light absorption
    Fig. 2. Schematic of underwater light absorption
    Contrast color-compensated image and original image. (a) Original image; (b) red channel; (c) green channel; (d) blue channel; (e) color compensated image based on bright channel; (f) color compensated red channel; (g) color compensated green channel; (h) color compensated blue channel
    Fig. 3. Contrast color-compensated image and original image. (a) Original image; (b) red channel; (c) green channel; (d) blue channel; (e) color compensated image based on bright channel; (f) color compensated red channel; (g) color compensated green channel; (h) color compensated blue channel
    Schematic of underwater optical imaging
    Fig. 4. Schematic of underwater optical imaging
    Adaptive contrast-stretched image. (a) Original image; (b) contrast-stretched image
    Fig. 5. Adaptive contrast-stretched image. (a) Original image; (b) contrast-stretched image
    Weight maps. (a) Color-compensated image (input1); (b) normalized weight maps for input1; (c) contrast-stretched image (input2); (d) normalized weight maps for input2
    Fig. 6. Weight maps. (a) Color-compensated image (input1); (b) normalized weight maps for input1; (c) contrast-stretched image (input2); (d) normalized weight maps for input2
    Color recovery test. (a) Standard color checker; (b) original image; (c) CALHE method; (d) DCP method; (e) MSRCR method; (f) proposed method
    Fig. 7. Color recovery test. (a) Standard color checker; (b) original image; (c) CALHE method; (d) DCP method; (e) MSRCR method; (f) proposed method
    Color distortion image test. (a) Original image; (b) CALHE method; (c) DCP method; (d) MSRCR method; (e) proposed method
    Fig. 8. Color distortion image test. (a) Original image; (b) CALHE method; (c) DCP method; (d) MSRCR method; (e) proposed method
    Hazed image test. (a) Group of original images; (b) CALHE method; (c) DCP method; (d) MSRCR method; (e) proposed method
    Fig. 9. Hazed image test. (a) Group of original images; (b) CALHE method; (c) DCP method; (d) MSRCR method; (e) proposed method
    Application test. (a) Feature point matching in the original image; (b) feature point matching in image processed using the proposed method
    Fig. 10. Application test. (a) Feature point matching in the original image; (b) feature point matching in image processed using the proposed method
    Image setCLAHEDCPMSRCRProposed method
    MSEPSNRMSEPSNRMSEPSNRMSEPSNR
    Image117334.75.7421286.117.03830025.73.35613756.76.746
    Image225656.74.039131.226.95040382.62.06910960.77.732
    Image310310.87.99810852.37.7765125.311.0345098.911.189
    Image411414.67.55611241.37.6237356.39.4642926.1813.468
    Image56096.610.2805987.610.3588335.18.9223926.312.191
    Image69868.38.1856853.79.7729747.08.2425490.110.735
    Image77516.79.37117752.85.6387423.79.4253174.013.115
    Table 1. Quality evaluation of underwater images
    Image setOriginal imageProposed method
    First group428
    Second group819
    Third group656
    Forth group717
    Table 2. Number of feature points matching contrast in original images and images processed using our method
    Chenggang Dai, Mingxing Lin, Zhen Wang, Dong Zhang, Zhiguang Guan. Color Compensation Based on Bright Channel and Fusion for Underwater Image Enhancement[J]. Acta Optica Sinica, 2018, 38(11): 1110003
    Download Citation