Real-Time Enhancement of Low-Light Images Using Generative Adversarial Networks (GANs)
DOI:
https://doi.org/10.51903/jtie.v4i1.279Keywords:
Low-Light Enhancement, GANs, Deep Learning, Image Processing, Real-Time ProcessingAbstract
Low-light image enhancement plays a crucial role in fields such as surveillance, photography, and medical imaging, where inadequate lighting significantly reduces image quality, leading to loss of detail and increased noise. Traditional enhancement methods, such as histogram equalization and Retinex, struggle to preserve fine details and often amplify noise, limiting their effectiveness in real-world applications. To address these issues, this study proposes a Generative Adversarial Networks (GANs)-based model to enhance low-light images in real-time while maintaining high visual fidelity. The model aims to improve contrast, reduce noise, and retain image structure more effectively than conventional methods. The proposed GAN model is trained using the LOL and SID datasets and evaluated using the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM). Experimental results show that the method achieves a PSNR of 28.4 dB and SSIM of 0.91, outperforming histogram equalization (PSNR: 18.5 dB, SSIM: 0.65) and Retinex (PSNR: 20.3 dB, SSIM: 0.72). Although the model operates in real-time, its inference time of 35.6 ms per image suggests further optimization to support edge computing applications. This study demonstrates that GAN-based enhancement significantly improves low-light images by preserving structural integrity while reducing noise. Future research should focus on optimizing the model for faster processing, experimenting with larger and more diverse datasets, and integrating the system into real-world applications such as automated surveillance and smart camera technologies.
References
Cao, P., Niu, Q., Zhu, Y., & Li, T. (2024). A Zero-Reference Low-Light Image-Enhancement Approach Based on Noise Estimation. Applied Sciences, 14(7), 2846. https://doi.org/10.3390/app14072846
Fan, Y., Wang, Y., Feng, K., Liu, Y., Jiang, Y., Xie, J., Niu, Y., & Wang, H. (2023). Laser Image Enhancement Algorithm Based on Improved EnlightenGAN. Electronics, 12(9), 1–13. https://doi.org/10.3390/electronics12092081
Farhadinia, B., Ahangari, M. R., Heydari, A., & Datta, A. (2024). A Generalized Optimization-Based Generative Adversarial Network. Expert Systems with Applications, 248, 123413. https://doi.org/10.1016/j.eswa.2024.123413
Guo, X., & Hu, Q. (2023). Low-light Image Enhancement via Breaking Down the Darkness. International Journal of Computer Vision, 131(1), 48–66. https://doi.org/10.1007/s11263-022-01667-9
Han, Y., Chen, X., Zhong, Y., Huang, Y., Li, Z., Han, P., Li, Q., & Yuan, Z. (2023). Low-Illumination Road Image Enhancement by Fusing Retinex Theory and Histogram Equalization. Electronics, 12(4), 990. https://doi.org/10.3390/electronics12040990
Hassan, N., Musa Miah, A. S., & Shin, J. (2024). Residual-Based Multi-Stage Deep Learning Framework for Computer-Aided Alzheimer’s Disease Detection. Journal of Imaging, 10(6), 141. https://doi.org/10.3390/jimaging10060141
Jenkins, J., & Roy, K. (2024). Exploring Deep Convolutional Generative Adversarial Networks (DCGAN) in Biometric Systems: A Survey Study. Discover Artificial Intelligence, 4(1), 42. https://doi.org/10.1007/s44163-024-00138-z
Ji, X., Guo, S., Zhang, H., & Xu, W. (2023). Non-Uniform-Illumination Image Enhancement Algorithm Based on Retinex Theory. Applied Sciences, 13(17), 9535. https://doi.org/10.3390/app13179535
Jiménez-Gaona, Y., Carrión-Figueroa, D., Lakshminarayanan, V., & José Rodríguez-Álvarez, M. (2024). Gan-Based Data Augmentation to Improve Breast Ultrasound and Mammography Mass Classification. Biomedical Signal Processing and Control, 94, 106255. https://doi.org/10.1016/j.bspc.2024.106255
Krstanović, L., Popović, B., Janev, M., & Brkljač, B. (2023). Feature Map Regularized CycleGAN for Domain Transfer. Mathematics, 11(2), 1–21. https://doi.org/10.3390/math11020372
Li, X., Yuan, Y., Liu, G., Han, Z., & Stouffs, R. (2024). A Predictive Model for Daylight Performance Based on Multimodal Generative Adversarial Networks at the Early Design Stage. Energy and Buildings, 305, 113876. https://doi.org/10.1016/j.enbuild.2023.113876
Liu, Z., Ma, C., She, W., & Xie, M. (2024). Biomedical Image Segmentation Using Denoising Diffusion Probabilistic Models: A Comprehensive Review and Analysis. Applied Sciences, 14(2), 632. https://doi.org/10.3390/app14020632
Nandhini Abirami, R., Durai Raj Vincent, P. M., Srinivasan, K., Tariq, U., & Chang, C. Y. (2021). Deep CNN and Deep GAN in Computational Visual Perception-Driven Image Analysis. Complexity, 2021(1), 5541134. https://doi.org/10.1155/2021/5541134
Porkodi, S. P., Sarada, V., Maik, V., & Gurushankar, K. (2023). Generic Image Application Using GANs (Generative Adversarial Networks): A Review. Evolving Systems, 14(5), 903–917. https://doi.org/10.1007/s12530-022-09464-y
Rasheed, M. T., Guo, G., Shi, D., Khan, H., & Cheng, X. (2022). An Empirical Study on Retinex Methods for Low-Light Image Enhancement. Remote Sensing, 14(18), 4608. https://doi.org/10.3390/rs14184608
Remtulla, R., Samet, A., Kulbay, M., Akdag, A., Hocini, A., Volniansky, A., Kahn Ali, S., & Qian, C. X. (2025). A Future Picture: A Review of Current Generative Adversarial Neural Networks in Vitreoretinal Pathologies and Their Future Potentials. Biomedicines, 13(2), 284. https://doi.org/10.3390/biomedicines13020284
Taassori, M. (2024). Enhanced Wavelet-Based Medical Image Denoising with Bayesian-Optimized Bilateral Filtering. Sensors, 24(21), 6849. https://doi.org/10.3390/s24216849
Tang, H., Zhu, H., Fei, L., Wang, T., Cao, Y., & Xie, C. (2023). Low-Illumination Image Enhancement Based on Deep Learning Techniques: A Brief Review. Photonics, 10(2), 198. https://doi.org/10.3390/photonics10020198
Tao, Y., & Muller, J. P. (2021). Super-Resolution Restoration of Spaceborne Ultra-High-Resolution Images Using the UCL Optigan System. Remote Sensing, 13(12), 2269. https://doi.org/10.3390/rs13122269
Taye, M. M. (2023). Theoretical Understanding of Convolutional Neural Network: Concepts, Architectures, Applications, Future Directions. Computation, 11(3), 52. https://doi.org/10.3390/computation11030052
Tian, Z., Qu, P., Li, J., Sun, Y., Li, G., Liang, Z., & Zhang, W. (2023). A Survey of Deep Learning-Based Low-Light Image Enhancement. Sensors, 23(18), 7763. https://doi.org/10.3390/s23187763
Tirel, L., Ali, A. M., & Hashim, H. A. (2024). Novel Hybrid Integrated Pix2Pix and WGAN Model with Gradient Penalty for Binary Images Denoising. Systems and Soft Computing, 6, 200122. https://doi.org/10.1016/j.sasc.2024.200122
Wang, J., Yang, Y., Chen, Y., & Han, Y. (2021). LighterGAN: An Illumination Enhancement Method for Urban UAV Imagery. Remote Sensing, 13(7), 1–20. https://doi.org/10.3390/rs13071371
Wu, Y., Liang, L., Ling, S., & Gao, Z. (2024). Hierarchical Patch Aggregation Transformer for Motion Deblurring. Neural Processing Letters, 56(2), 1–22. https://doi.org/10.1007/s11063-024-11594-0
Xue, M., He, Y., Xie, P., He, Z., & Feng, X. (2023). LAE-GAN: A Novel Cloud-Based Low-Light Attention Enhancement Generative Adversarial Network for Unpaired Text Images. Journal of Cloud Computing, 12(1), 160. https://doi.org/10.1186/s13677-023-00533-4
Zhang, F., Liu, X., Gao, C., & Sang, N. (2024). Color and Luminance Separated Enhancement for Low-Light Images with Brightness Guidance. Sensors, 24(9), 2711. https://doi.org/10.3390/s24092711
Zhang, J., Ji, R., Wang, J., Sun, H., & Ju, M. (2023). DEGAN: Decompose-Enhance-GAN Network for Simultaneous Low-Light Image Lightening and Denoising. Electronics, 12(14), 3038. https://doi.org/10.3390/electronics12143038
Zhu, L., Shen, K., Wang, G., Hao, Y., Zheng, L., & Lu, Y. (2025). DUWS Net: Wavelet-Based Dual U-Shaped Spatial-Frequency Fusion Transformer Network for Medical Image Segmentation. Journal of Visual Communication and Image Representation, 108, 104428. https://doi.org/10.1016/j.jvcir.2025.104428
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Journal of Technology Informatics and Engineering
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
