Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models
Satellite imagery plays a pivotal role in environmental monitoring, urban planning, and national security. However, spatial resolution limitations of current satellite sensors restrict the clarity and usability of captured images. This study introduces a novel transformer-based deep-learning model t...
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/10829708/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1825206999790387200 |
|---|---|
| author | Alireza Sharifi Mohammad Mahdi Safari |
| author_facet | Alireza Sharifi Mohammad Mahdi Safari |
| author_sort | Alireza Sharifi |
| collection | DOAJ |
| description | Satellite imagery plays a pivotal role in environmental monitoring, urban planning, and national security. However, spatial resolution limitations of current satellite sensors restrict the clarity and usability of captured images. This study introduces a novel transformer-based deep-learning model to enhance the spatial resolution of Sentinel-2 images. The proposed architecture leverages multihead attention and integrated spatial and channel attention mechanisms to effectively extract and reconstruct fine details from low-resolution inputs. The model's performance was evaluated on the Sentinel-2 dataset, along with benchmark datasets (AID and UC-Merced), and compared against state-of-the-art methods, including ResNet, Swin Transformer, and ViT. Experimental results demonstrate superior performance, achieving a peak signal-to-noise ratio (PSNR) of 33.52 dB, structural similarity index (SSIM) of 0.862, and signal-to-reconstruction error ratio (SRE) of 36.7 dB on Sentinel-2 RGB bands. The proposed method outperforms state-of-the-art approaches, including ResNet, Swin Transformer, and ViT, on benchmark datasets (Sentinel-2, AID, and UC-Merced). The results demonstrate that the proposed method achieves superior performance in terms of PSNR, SSIM, and SRE metrics, highlighting its effectiveness in revealing finer spatial details and improving image quality for practical remote sensing applications. |
| format | Article |
| id | doaj-art-1048cb59268f49d295ff817c891fca2f |
| 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-1048cb59268f49d295ff817c891fca2f2025-02-07T00:00:33ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01184805482010.1109/JSTARS.2025.352626010829708Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning ModelsAlireza Sharifi0https://orcid.org/0000-0001-7110-7516Mohammad Mahdi Safari1https://orcid.org/0009-0009-5240-8466Department of Surveying Engineering, Faculty of Civil, Water and Environmental Engineering, Shahid Beheshti University, Tehran, IranDepartment of Geoinformatics Engineering, Politecnico di Milano, Milano, ItalySatellite imagery plays a pivotal role in environmental monitoring, urban planning, and national security. However, spatial resolution limitations of current satellite sensors restrict the clarity and usability of captured images. This study introduces a novel transformer-based deep-learning model to enhance the spatial resolution of Sentinel-2 images. The proposed architecture leverages multihead attention and integrated spatial and channel attention mechanisms to effectively extract and reconstruct fine details from low-resolution inputs. The model's performance was evaluated on the Sentinel-2 dataset, along with benchmark datasets (AID and UC-Merced), and compared against state-of-the-art methods, including ResNet, Swin Transformer, and ViT. Experimental results demonstrate superior performance, achieving a peak signal-to-noise ratio (PSNR) of 33.52 dB, structural similarity index (SSIM) of 0.862, and signal-to-reconstruction error ratio (SRE) of 36.7 dB on Sentinel-2 RGB bands. The proposed method outperforms state-of-the-art approaches, including ResNet, Swin Transformer, and ViT, on benchmark datasets (Sentinel-2, AID, and UC-Merced). The results demonstrate that the proposed method achieves superior performance in terms of PSNR, SSIM, and SRE metrics, highlighting its effectiveness in revealing finer spatial details and improving image quality for practical remote sensing applications.https://ieeexplore.ieee.org/document/10829708/Deep learningsatellite imagessuper-resolutiontransformer |
| spellingShingle | Alireza Sharifi Mohammad Mahdi Safari Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Deep learning satellite images super-resolution transformer |
| title | Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models |
| title_full | Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models |
| title_fullStr | Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models |
| title_full_unstemmed | Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models |
| title_short | Enhancing the Spatial Resolution of Sentinel-2 Images Through Super-Resolution Using Transformer-Based Deep-Learning Models |
| title_sort | enhancing the spatial resolution of sentinel 2 images through super resolution using transformer based deep learning models |
| topic | Deep learning satellite images super-resolution transformer |
| url | https://ieeexplore.ieee.org/document/10829708/ |
| work_keys_str_mv | AT alirezasharifi enhancingthespatialresolutionofsentinel2imagesthroughsuperresolutionusingtransformerbaseddeeplearningmodels AT mohammadmahdisafari enhancingthespatialresolutionofsentinel2imagesthroughsuperresolutionusingtransformerbaseddeeplearningmodels |