Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review
As a critical preprocessing technique, low-illumination image enhancement has a wide range of practical applications. It aims to improve the visual perception of a given image captured without sufficient illumination. Conventional low-illumination image enhancement methods are typically implemented...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Photonics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2304-6732/10/2/198 |
| _version_ | 1797618668094357504 |
|---|---|
| author | Hao Tang Hongyu Zhu Linfeng Fei Tingwei Wang Yichao Cao Chao Xie |
| author_facet | Hao Tang Hongyu Zhu Linfeng Fei Tingwei Wang Yichao Cao Chao Xie |
| author_sort | Hao Tang |
| collection | DOAJ |
| description | As a critical preprocessing technique, low-illumination image enhancement has a wide range of practical applications. It aims to improve the visual perception of a given image captured without sufficient illumination. Conventional low-illumination image enhancement methods are typically implemented by improving image brightness, enhancing image contrast, and suppressing image noise simultaneously. Nevertheless, recent advances in this area are dominated by deep-learning-based solutions, and consequently, various deep neural networks have been proposed and applied to this field. Therefore, this paper briefly reviews the latest low-illumination image enhancement, ranging from its related algorithms to its unsolved open issues. Specifically, current low-illumination image enhancement methods based on deep learning are first sorted out and divided into four categories: supervised learning methods, unsupervised learning methods, semi-supervised learning methods, and zero-shot learning methods. Then, existing low-light image datasets are summarized and analyzed. In addition, various quality assessment indices for low-light image enhancement are introduced in detail. We also compare 14 representative algorithms in terms of both objective evaluation and subjective evaluation. Finally, the future development trend of low-illumination image enhancement and its open issues are summarized and prospected. |
| first_indexed | 2024-03-11T08:15:45Z |
| format | Article |
| id | doaj.art-5ed039c26a414b57aaba3d69cc744c09 |
| institution | Directory Open Access Journal |
| issn | 2304-6732 |
| language | English |
| last_indexed | 2024-03-11T08:15:45Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Photonics |
| spelling | doaj.art-5ed039c26a414b57aaba3d69cc744c092023-11-16T22:45:42ZengMDPI AGPhotonics2304-67322023-02-0110219810.3390/photonics10020198Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief ReviewHao Tang0Hongyu Zhu1Linfeng Fei2Tingwei Wang3Yichao Cao4Chao Xie5College of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, ChinaCollege of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, ChinaSchool of Automation, Southeast University, Nanjing 210096, ChinaCollege of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, ChinaAs a critical preprocessing technique, low-illumination image enhancement has a wide range of practical applications. It aims to improve the visual perception of a given image captured without sufficient illumination. Conventional low-illumination image enhancement methods are typically implemented by improving image brightness, enhancing image contrast, and suppressing image noise simultaneously. Nevertheless, recent advances in this area are dominated by deep-learning-based solutions, and consequently, various deep neural networks have been proposed and applied to this field. Therefore, this paper briefly reviews the latest low-illumination image enhancement, ranging from its related algorithms to its unsolved open issues. Specifically, current low-illumination image enhancement methods based on deep learning are first sorted out and divided into four categories: supervised learning methods, unsupervised learning methods, semi-supervised learning methods, and zero-shot learning methods. Then, existing low-light image datasets are summarized and analyzed. In addition, various quality assessment indices for low-light image enhancement are introduced in detail. We also compare 14 representative algorithms in terms of both objective evaluation and subjective evaluation. Finally, the future development trend of low-illumination image enhancement and its open issues are summarized and prospected.https://www.mdpi.com/2304-6732/10/2/198deep learninglow-illumination image enhancementRetinex theoryquality evaluation indeximage dataset |
| spellingShingle | Hao Tang Hongyu Zhu Linfeng Fei Tingwei Wang Yichao Cao Chao Xie Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review Photonics deep learning low-illumination image enhancement Retinex theory quality evaluation index image dataset |
| title | Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review |
| title_full | Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review |
| title_fullStr | Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review |
| title_full_unstemmed | Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review |
| title_short | Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review |
| title_sort | low illumination image enhancement based on deep learning techniques a brief review |
| topic | deep learning low-illumination image enhancement Retinex theory quality evaluation index image dataset |
| url | https://www.mdpi.com/2304-6732/10/2/198 |
| work_keys_str_mv | AT haotang lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview AT hongyuzhu lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview AT linfengfei lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview AT tingweiwang lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview AT yichaocao lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview AT chaoxie lowilluminationimageenhancementbasedondeeplearningtechniquesabriefreview |