Acta Optica Sinica, Volume. 38, Issue 4, 0410002(2018)
Study On Image Dehazing Methods Based On Dark Channel Prior
Fig. 1. (a) Original image and haze removal results after select different window sizes of (b) 10, (c) 50, and (d) 100
Fig. 2. (a) Original image and haze removal results after select different ω values of (b) 0.5, (c) 0.75, and (d) 0.95
Fig. 3. Discovery of "halo" phenomenon. (a) Original image; (b) dark channel of original image; (c) transmission ratio image; (d) haze removal result
Fig. 5. (a) Hazy image and transmission ratio images with different filtering parameter r values of (b) 1, (c) 37, and (d) 100
Fig. 6. (a) Hazy image and haze removal results with different filtering parameter r values of (b) 1, (c) 37, and (d) 100
Fig. 7. (a) Hazy image and transmission ratio images with different adjustment parameter ε values of (b) 10-1, (c) 10-3, and (d) 10-6
Fig. 8. Comparison of (a) hazy image and (b) haze removal result without special sky treatment
Fig. 9. Haze removal results after sky treatment with different threshold A0 values of (a) 80, (b) 150, (c) 200, (d) 210, (e) 220,; (f) 230, (g) 240, and (h) 255
Fig. 10. Comparison of images with different interpolation methods. (a) Hazy image; (b) nearest neighbour interpolation; (c) bilinear interpolation; (d) cubic convolution interpolation
Fig. 11. Comparison of transmission ratio images (a) without and (b) with interpolation method when N=5
Fig. 12. (a) Original images and haze removal results with different algorithms of (b) histogram equalization, (c) Retinex, (d) He algorithm, (e) He algorithm combined with guided-filter, and (f) proposed algorithm
|
Get Citation
Copy Citation Text
Han Guo, Xiaoting Xu, Bo Li. Study On Image Dehazing Methods Based On Dark Channel Prior[J]. Acta Optica Sinica, 2018, 38(4): 0410002
Category: Image Processing
Received: Sep. 26, 2017
Accepted: --
Published Online: Jul. 10, 2018
The Author Email: Li Bo (libo@zjut.edu.cn)