Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm
Researchers working on image processing have had a hard time handling low-light images due to their low contrast, noise, and brightness. This paper presents an improved method that uses the Retinex theory to enhance low-light images, with a network model mainly composed of a Decom-Net and an Enhance...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/14/8148 |
| _version_ | 1797590440329871360 |
|---|---|
| author | Jiarui Wang Hanjia Wang Yu Sun Jie Yang |
| author_facet | Jiarui Wang Hanjia Wang Yu Sun Jie Yang |
| author_sort | Jiarui Wang |
| collection | DOAJ |
| description | Researchers working on image processing have had a hard time handling low-light images due to their low contrast, noise, and brightness. This paper presents an improved method that uses the Retinex theory to enhance low-light images, with a network model mainly composed of a Decom-Net and an Enhance-Net. Residual connectivity is fully utilized in both the Decom-Net and Enhance-Net to reduce the possible loss of image details. Additionally, Enhance-Net introduces a positional pixel attention mechanism that directly incorporates the global information of the image. Specifically, Decom-Net serves to decompose the low-light image into illumination and reflection maps, and Enhance-Net serves to increase the brightness of the illumination map. Finally, via adaptive image fusion, the reflectance map and the enhanced illuminance map are fused to obtain the final enhanced image. Experiments show better results in terms of both subjective visual aspects and objective evaluation indicators. Compared to RetinexNet, the proposed method shows improvements in the full-reference evaluation metrics, including a 4.6% improvement in PSNR, a 1.8% improvement in SSIM, and a 10.8% improvement in LPIPS. Additionally, it achieved an average improvement of 17.3% in the no-reference evaluation metric NIQE. |
| first_indexed | 2024-03-11T01:20:31Z |
| format | Article |
| id | doaj.art-4de7ae55ee134dec86296ae840a5955d |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-11T01:20:31Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-4de7ae55ee134dec86296ae840a5955d2023-11-18T18:08:56ZengMDPI AGApplied Sciences2076-34172023-07-011314814810.3390/app13148148Improved Retinex-Theory-Based Low-Light Image Enhancement AlgorithmJiarui Wang0Hanjia Wang1Yu Sun2Jie Yang3Dalian Naval Academy, Dalian 116013, ChinaDalian Naval Academy, Dalian 116013, ChinaDalian Naval Academy, Dalian 116013, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaResearchers working on image processing have had a hard time handling low-light images due to their low contrast, noise, and brightness. This paper presents an improved method that uses the Retinex theory to enhance low-light images, with a network model mainly composed of a Decom-Net and an Enhance-Net. Residual connectivity is fully utilized in both the Decom-Net and Enhance-Net to reduce the possible loss of image details. Additionally, Enhance-Net introduces a positional pixel attention mechanism that directly incorporates the global information of the image. Specifically, Decom-Net serves to decompose the low-light image into illumination and reflection maps, and Enhance-Net serves to increase the brightness of the illumination map. Finally, via adaptive image fusion, the reflectance map and the enhanced illuminance map are fused to obtain the final enhanced image. Experiments show better results in terms of both subjective visual aspects and objective evaluation indicators. Compared to RetinexNet, the proposed method shows improvements in the full-reference evaluation metrics, including a 4.6% improvement in PSNR, a 1.8% improvement in SSIM, and a 10.8% improvement in LPIPS. Additionally, it achieved an average improvement of 17.3% in the no-reference evaluation metric NIQE.https://www.mdpi.com/2076-3417/13/14/8148image processinglow-light image enhancementRetinex theorylow/normal-light image |
| spellingShingle | Jiarui Wang Hanjia Wang Yu Sun Jie Yang Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm Applied Sciences image processing low-light image enhancement Retinex theory low/normal-light image |
| title | Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm |
| title_full | Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm |
| title_fullStr | Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm |
| title_full_unstemmed | Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm |
| title_short | Improved Retinex-Theory-Based Low-Light Image Enhancement Algorithm |
| title_sort | improved retinex theory based low light image enhancement algorithm |
| topic | image processing low-light image enhancement Retinex theory low/normal-light image |
| url | https://www.mdpi.com/2076-3417/13/14/8148 |
| work_keys_str_mv | AT jiaruiwang improvedretinextheorybasedlowlightimageenhancementalgorithm AT hanjiawang improvedretinextheorybasedlowlightimageenhancementalgorithm AT yusun improvedretinextheorybasedlowlightimageenhancementalgorithm AT jieyang improvedretinextheorybasedlowlightimageenhancementalgorithm |