Researching | Low-light color image enhancement algorithm based on conditional generative adversarial network

Infrared and Laser Engineering, Volume. 53, Issue 11, 20240238(2024)

Low-light color image enhancement algorithm based on conditional generative adversarial network

Jue WANG1... Minxuan HONG1, Yetong XIA1, Xiuyu XU1,2,3, Xiaofang KONG4, and Minjie WAN1,23 |Show fewer author(s)

Author Affiliations

¹School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China

²Jiangsu Key Laboratory of Spectral Imaging & Intelligent Sense, Nanjing University of Science and Technology, Nanjing 210094, China

³Advanced Interdisciplinary Research Center for Optics, Nanjing University of Science and Technology, Nanjing 210094, China

⁴National Key Laboratory of Transient Physics, Nanjing University of Science and Technology, Nanjing 210094, China

show less

Abstract Get PDF(in Chinese)

Figures&Tables (15)

References (35)

Paper Information

Figures & Tables(15)

Fig. 1. Structure of CGAN

Download full size

View in Article

Fig. 2. Structure of the generator network

Download full size

View in Article

Fig. 3. Structure of the discriminator network

Download full size

View in Article

Fig. 4. Structure of the RRDB network

Download full size

View in Article

Fig. 5. Structure of the Dense Block network

Download full size

View in Article

Fig. 6. Channel attention module

Download full size

View in Article

Fig. 7. Spatial attention module

Download full size

View in Article

Fig. 8. CBAM diagram

Download full size

View in Article

Fig. 9. SK-Net module

Download full size

View in Article

Fig. 10. Comparison of different algorithms

Download full size

View in Article

Fig. 11. Outdoor scene image test results

Download full size

View in Article

Table 1. Evaluating indicator

View in Article

Table 1. Evaluating indicator

Algorithm	SSIM	PSNR	CD
Input image	0.2291	6.5025	4381.3892
JED	0.6352	19.5944	1116.9907
KinD_New	0.8399	21.9442	665.6067
Retinex-Net	0.6386	18.0844	850.4218
SNR	0.8393	27.0077	487.4247
URetinex-Net	0.8312	22.6170	745.8164
Ours	0.8478	28.5475	438.3771

Table 2. Ablation experiment

View in Article

Table 2. Ablation experiment

Content_loss	Prewitt_loss	YUV_loss	SSIM	PSNR	CD
√	√	√	0.8478	28.5475	438.3771
-	√	√	0.0187	6.5450	4998.8540
-	-	√	0.0187	6.5452	4998.8540
-	√	-	0.1646	7.9943	1989.0770
√	-	√	0.8373	28.0317	478.8337
√	√	-	0.1859	8.2461	1621.2470
√	-	-	0.4654	22.3765	615.3432

Table 3. Processing speed
View table
View in Article
Table 3. Processing speed
Algorithm Frame rate/frame·s^-1
JED 0.55
KinD_New 23.86
Retinex-Net 20.33
SNR 19.53
URetinex-Net 15.58
Ours 11.13

Table 4. Outdoor scene evaluation index results
View table
View in Article
Table 4. Outdoor scene evaluation index results
Index SSIM PSNR CD
Ours 0.7168 26.0069 441.7949

Tools

Get Citation

Copy Citation Text

Jue WANG, Minxuan HONG, Yetong XIA, Xiuyu XU, Xiaofang KONG, Minjie WAN. Low-light color image enhancement algorithm based on conditional generative adversarial network[J]. Infrared and Laser Engineering, 2024, 53(11): 20240238

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: 图像处理

Received: Jul. 19, 2024

Accepted: --

Published Online: Dec. 13, 2024

The Author Email:

DOI:10.3788/IRLA20240238

Topics

laser devices and laser physics

Lasers and Laser Optics

laser manufacturing

Instrumentation, Measurement and Metrology