Adaptive Image Dehazing Algorithm Based on Dark Channel Prior Theory

Jiyu Xiao; Yunmeng Liu; Shizhao Li; Lei Ding

doi:10.3788/LOP242440

Laser & Optoelectronics Progress, Volume. 62, Issue 14, 1401001(2025)

Adaptive Image Dehazing Algorithm Based on Dark Channel Prior Theory

Jiyu Xiao^1,2,3, Yunmeng Liu^1,3、**, Shizhao Li^1,3, and Lei Ding^1,2,3、*

¹Key Laboratory of Infrared System Detection and Imaging Technology, Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai 200083, China

²School of Information Science and Technology, ShanghaiTech University, Shanghai 201210, China

³Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai 200083, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. Grayscale value distribution diagrams of the original and histogram equalized images. (a) Original image; (b) grayscale value distribution diagram of the original image; (c) grayscale value distribution map of the histogram equalized image

Download full size

Fig. 2. Original and CLAHE processed images. (a)‒(c) Original images; (d)‒(f) the images after CLAHE processing

Download full size

Fig. 3. Original and gamma corrected images. (a)‒(c) Original images; (d)‒(f) gamma corrected images

Download full size

Fig. 4. The images processed by different algorithms. (a) (d) The images processed by CLAHE algorithm; (b)(e) the images processed by improved DCP algorithm; (c) (f) fusion images

Download full size

Fig. 5. Improved dark channel dehazing algorithm flowchart

Download full size

Fig. 6. The relationship between peak signal-to-noise-ratio of the dehazed images, A and ω

Download full size

Fig. 7. Flow diagram showing improved calculation of atmospheric light value and transmittance

Download full size

Fig. 8. Original images and the images processed by traditional and improved dark channel dehazing algorithms

Download full size

Fig. 9. Original images and the images processed by traditional dark channel dehazing algorithm, Retinex dehazing algorithm, and improved dark channel dehazing algorithm

Download full size

Fig. 10. Original images and the images processed by traditional and improved dark channel dehazing algorithms

Download full size

Table 1. Comparison of PSNR, MSE, and SSIM for different algorithms

View table

Table 1. Comparison of PSNR, MSE, and SSIM for different algorithms

Image	Original algorithm			Retinex algorithm			Improved algorithm
Image	PSNR	MSE	SSIM	PSNR	MSE	SSIM	PSNR	MSE	SSIM
No.1	56.9341	0.1317	0.7703	58.3229	0.0957	0.4886	84.8869	0.0002	0.9856
No.2	59.2777	0.0768	0.7283	59.4117	0.0745	0.5035	79.7892	0.0007	0.9893
No.3	56.2920	0.1527	0.7406	55.5184	0.1825	0.5079	83.8223	0.0003	0.9864

Table 2. Comparison of U, $\bar{r}$ , UISM for different algorithms

View table

Table 2. Comparison of U, $\bar{r}$ , UISM for different algorithms

Image	Original algorithm			Retinex algorithm			Improved algorithm
Image	U	$\bar{r}$	UISM	U	$\bar{r}$	UISM	U	$\bar{r}$	UISM
No.1	7.5348	0.8896	0.5747	5.8612	0.8979	0.6223	7.6269	0.9228	0.8164
No.2	7.2767	0.6176	1.1100	7.0012	0.6778	1.1133	7.4285	0.7250	1.3311
No.3	7.3685	0.8345	0.6137	4.7408	0.8544	0.7762	7.6769	0.8967	0.8532

Table 3. Quantitative indexes comparison between traditional and improved dark channel dehazing algorithms

View table

Table 3. Quantitative indexes comparison between traditional and improved dark channel dehazing algorithms

Algorithm	Image	PSNR	MSE	SSIM	U	$\bar{r}$	UISM
Traditional	No.1	59.2634	0.0770	0.7578	7.0951	0.9484	0.6099
Traditional	No.2	62.2001	0.0392	0.7164	6.7713	0.9712	0.5497
Improved	No.1	68.4852	0.0092	0.8889	7.5165	0.9998	0.6915
Improved	No.2	64.9691	0.0207	0.8166	7.5128	0.9931	0.5995

Tools

Get Citation

Copy Citation Text

Jiyu Xiao, Yunmeng Liu, Shizhao Li, Lei Ding. Adaptive Image Dehazing Algorithm Based on Dark Channel Prior Theory[J]. Laser & Optoelectronics Progress, 2025, 62(14): 1401001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites