Image Defogging Based on Combination of Image Bright and Dark Channels

Huibin Lu; Yanfang Zhao; Yongjie Zhao; Shuhuan Wen; Jinrong Ma; Hak Keung Lam; Hongbin Wang

doi:10.3788/AOS201838.1115004

Acta Optica Sinica, Volume. 38, Issue 11, 1115004(2018)

Image Defogging Based on Combination of Image Bright and Dark Channels

Huibin Lu^1、*, Yanfang Zhao², Yongjie Zhao², Shuhuan Wen^2、*, Jinrong Ma², Hak Keung Lam, and Hongbin Wang²

¹ Key Lab of Information Transmission and Signal Processing of Hebei Province, School of Information Science and Engineering, Yanshan University, Qinhuangdao, Hebei 0 66004, China

² Key Lab of Industrial Computer Control Engineering of Hebei Province, Yanshan University, Qinhuangdao, Hebei 0 66004, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(8)

Fig. 1. Different light intensity in different regions

Download full size

Fig. 2. Flow chart of image defogging algorithm

Download full size

Fig. 3. (a) Foggy images; (b) dark channel images; (c) bright channel images

Download full size

Fig. 4. Transmittance images and restored images. (a) Initial estimated transmittance images; (b) optimized transmittance images; (c) (d) recovered images from Fig. (a) and (b) respectively

Download full size

Fig. 5. Foggy images and defogging results of different algorithm. (a) Foggy images; (b) multi-scale Retinex method; (c) method in Ref. [10]; (d) proposed method

Download full size

Fig. 6. Foggy images and defogging results of different algorithm. (a) Foggy images; (b) multi-scale Retinex method; (c) method in Ref. [10]; (d) proposed method

Download full size

Table 1. Objective evaluation standards of different algorithms in Fig. 5
View table
Table 1. Objective evaluation standards of different algorithms in Fig. 5
Index Original image Multi-scale Retinex Algorithm in Ref. [10] Proposed algorithm
Average 103.53 129.01 112.35 138.80
Variance 39.93 55.23 49.35 58.55
Entropy 7.11 7.14 7.07 7.21
MSE - 100 116 95
PSNR - 65.83 63.28 61.55

Table 2. Objective evaluation standards of different algorithms in Fig. 6
View table
Table 2. Objective evaluation standards of different algorithms in Fig. 6
Index Original image Multi-scale Retinex Algorithm in Ref. [10] Proposed algorithm
Average 116.80 121.19 97.22 131.21
Variance 40.76 48.55 40.61 50.75
Entropy 7.18 7.15 7.05 7.25
MSE - 97 102 91
PSNR - 65.08 65.72 63.71

Tools

Get Citation

Copy Citation Text

Huibin Lu, Yanfang Zhao, Yongjie Zhao, Shuhuan Wen, Jinrong Ma, Hak Keung Lam, Hongbin Wang. Image Defogging Based on Combination of Image Bright and Dark Channels[J]. Acta Optica Sinica, 2018, 38(11): 1115004

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites