Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation
Image super-resolution (SR) aims to enhance the spatial resolution of images and overcome the hardware limitations of imaging systems. While deep-learning networks have significantly improved SR performance, obtaining paired low-resolution (LR) and high-resolution (HR) images for supervised learning...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10844307/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832542571178491904 |
|---|---|
| author | Minkyung Chung Yongil Kim |
| author_facet | Minkyung Chung Yongil Kim |
| author_sort | Minkyung Chung |
| collection | DOAJ |
| description | Image super-resolution (SR) aims to enhance the spatial resolution of images and overcome the hardware limitations of imaging systems. While deep-learning networks have significantly improved SR performance, obtaining paired low-resolution (LR) and high-resolution (HR) images for supervised learning remains challenging in real-world scenarios. In this article, we propose a novel unsupervised image super-resolution model for real-world remote sensing images, specifically focusing on HR satellite imagery. Our model, the bicubic-downsampled LR image-guided generative adversarial network for unsupervised learning (BLG-GAN-U), divides the SR process into two stages: LR image domain translation and image super-resolution. To implement this division, the model integrates omnidirectional real-to-synthetic domain translation with training strategies such as frequency separation and guided filtering. The model was evaluated through comparative analyses and ablation studies using real-world LR–HR datasets from WorldView-3 HR satellite imagery. The experimental results demonstrate that BLG-GAN-U effectively generates high-quality SR images with excellent perceptual quality and reasonable image fidelity, even with a relatively smaller network capacity. |
| format | Article |
| id | doaj-art-38594f5e5fc147d6a2eb44c19a3e9a4b |
| institution | Kabale University |
| issn | 1939-1404 2151-1535 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj-art-38594f5e5fc147d6a2eb44c19a3e9a4b2025-02-04T00:00:17ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01184427444510.1109/JSTARS.2025.353095910844307Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain TranslationMinkyung Chung0https://orcid.org/0000-0002-1266-5566Yongil Kim1https://orcid.org/0000-0003-0541-8986Department of Civil and Environmental Engineering, Seoul National University, Seoul, South KoreaDepartment of Civil and Environmental Engineering, Seoul National University, Seoul, South KoreaImage super-resolution (SR) aims to enhance the spatial resolution of images and overcome the hardware limitations of imaging systems. While deep-learning networks have significantly improved SR performance, obtaining paired low-resolution (LR) and high-resolution (HR) images for supervised learning remains challenging in real-world scenarios. In this article, we propose a novel unsupervised image super-resolution model for real-world remote sensing images, specifically focusing on HR satellite imagery. Our model, the bicubic-downsampled LR image-guided generative adversarial network for unsupervised learning (BLG-GAN-U), divides the SR process into two stages: LR image domain translation and image super-resolution. To implement this division, the model integrates omnidirectional real-to-synthetic domain translation with training strategies such as frequency separation and guided filtering. The model was evaluated through comparative analyses and ablation studies using real-world LR–HR datasets from WorldView-3 HR satellite imagery. The experimental results demonstrate that BLG-GAN-U effectively generates high-quality SR images with excellent perceptual quality and reasonable image fidelity, even with a relatively smaller network capacity.https://ieeexplore.ieee.org/document/10844307/Domain translationhigh-resolution (HR) satellite imageryimage super-resolution (SR)unsupervised learning |
| spellingShingle | Minkyung Chung Yongil Kim Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Domain translation high-resolution (HR) satellite imagery image super-resolution (SR) unsupervised learning |
| title | Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation |
| title_full | Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation |
| title_fullStr | Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation |
| title_full_unstemmed | Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation |
| title_short | Unsupervised Image Super-Resolution for High-Resolution Satellite Imagery via Omnidirectional Real-to-Synthetic Domain Translation |
| title_sort | unsupervised image super resolution for high resolution satellite imagery via omnidirectional real to synthetic domain translation |
| topic | Domain translation high-resolution (HR) satellite imagery image super-resolution (SR) unsupervised learning |
| url | https://ieeexplore.ieee.org/document/10844307/ |
| work_keys_str_mv | AT minkyungchung unsupervisedimagesuperresolutionforhighresolutionsatelliteimageryviaomnidirectionalrealtosyntheticdomaintranslation AT yongilkim unsupervisedimagesuperresolutionforhighresolutionsatelliteimageryviaomnidirectionalrealtosyntheticdomaintranslation |