A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing
Abstract Object detection in remote sensing imagery presents challenges due to low resolution, complex backgrounds, occlusions, and scale variations, which are critical in disaster response, environmental monitoring, and surveillance. This study proposes a robust object detection framework integrati...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-05-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-01476-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850273460972945408 |
|---|---|
| author | Muhammad Asif Mohammad Abrar Faizan Ullah Abdu Salam Farhan Amin Isabel de la Torre Mónica Gracia Villar Helena Garay Gyu Sang Choi |
| author_facet | Muhammad Asif Mohammad Abrar Faizan Ullah Abdu Salam Farhan Amin Isabel de la Torre Mónica Gracia Villar Helena Garay Gyu Sang Choi |
| author_sort | Muhammad Asif |
| collection | DOAJ |
| description | Abstract Object detection in remote sensing imagery presents challenges due to low resolution, complex backgrounds, occlusions, and scale variations, which are critical in disaster response, environmental monitoring, and surveillance. This study proposes a robust object detection framework integrating super-resolution techniques with advanced feature extraction algorithms for remote sensing images. The hybrid model combines Advanced StyleGAN and Swin Transformer. Advanced StyleGAN enhances image resolution, facilitating the detection of small and occluded objects, while Swin Transformer employs hierarchical attention mechanisms for effective feature extraction. Preprocessing techniques, including data augmentation, are incorporated to improve the diversity and accuracy of the training dataset. Evaluation on datasets such as VEDAI-VISIBLE and VEDAI-IR demonstrated exceptional performance, achieving an mAP@0.5 of 97.2%, mAP@0.5:0.95 of 72.8%, and F1-Score of 0.93, with an inference time of 42 ms. The framework maintained robustness under challenging conditions, such as low light and fog, outperforming YOLOv9-S, YOLOv9-E, and DCNN-based methods. Furthermore, it surpassed state-of-the-art models on RSOD and NWPU VHR-10 datasets, achieving superior detection accuracy and robustness. This framework offers a significant advancement in remote sensing object detection, providing an effective solution for complex scenarios. Future work may focus on optimizing computational efficiency and expanding the framework to multimodal or dynamic object detection tasks. |
| format | Article |
| id | doaj-art-e59e8028fc8e44d9a97b15f6ffa8819c |
| institution | OA Journals |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-e59e8028fc8e44d9a97b15f6ffa8819c2025-08-20T01:51:28ZengNature PortfolioScientific Reports2045-23222025-05-0115111710.1038/s41598-025-01476-3A novel hybrid deep learning approach for super-resolution and objects detection in remote sensingMuhammad Asif0Mohammad Abrar1Faizan Ullah2Abdu Salam3Farhan Amin4Isabel de la Torre5Mónica Gracia Villar6Helena Garay7Gyu Sang Choi8Department of Computer Science, Bacha Khan UniversityFaculty of Computer Studies, Arab Open UniversityDepartment of Computer Science, Bacha Khan UniversityDepartment of Computer Science, Abdul Wali Khan University MardanSchool of Computer Science and Engineering, Yeungnam UniversityDepartment of Signal Theory and Communications, University of ValladolidUniversidad Europea del AtlánticoUniversidad Europea del AtlánticoSchool of Computer Science and Engineering, Yeungnam UniversityAbstract Object detection in remote sensing imagery presents challenges due to low resolution, complex backgrounds, occlusions, and scale variations, which are critical in disaster response, environmental monitoring, and surveillance. This study proposes a robust object detection framework integrating super-resolution techniques with advanced feature extraction algorithms for remote sensing images. The hybrid model combines Advanced StyleGAN and Swin Transformer. Advanced StyleGAN enhances image resolution, facilitating the detection of small and occluded objects, while Swin Transformer employs hierarchical attention mechanisms for effective feature extraction. Preprocessing techniques, including data augmentation, are incorporated to improve the diversity and accuracy of the training dataset. Evaluation on datasets such as VEDAI-VISIBLE and VEDAI-IR demonstrated exceptional performance, achieving an mAP@0.5 of 97.2%, mAP@0.5:0.95 of 72.8%, and F1-Score of 0.93, with an inference time of 42 ms. The framework maintained robustness under challenging conditions, such as low light and fog, outperforming YOLOv9-S, YOLOv9-E, and DCNN-based methods. Furthermore, it surpassed state-of-the-art models on RSOD and NWPU VHR-10 datasets, achieving superior detection accuracy and robustness. This framework offers a significant advancement in remote sensing object detection, providing an effective solution for complex scenarios. Future work may focus on optimizing computational efficiency and expanding the framework to multimodal or dynamic object detection tasks.https://doi.org/10.1038/s41598-025-01476-3Super-resolutionHigh resolutionObject detectionDeep learningRemote sensing |
| spellingShingle | Muhammad Asif Mohammad Abrar Faizan Ullah Abdu Salam Farhan Amin Isabel de la Torre Mónica Gracia Villar Helena Garay Gyu Sang Choi A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing Scientific Reports Super-resolution High resolution Object detection Deep learning Remote sensing |
| title | A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing |
| title_full | A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing |
| title_fullStr | A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing |
| title_full_unstemmed | A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing |
| title_short | A novel hybrid deep learning approach for super-resolution and objects detection in remote sensing |
| title_sort | novel hybrid deep learning approach for super resolution and objects detection in remote sensing |
| topic | Super-resolution High resolution Object detection Deep learning Remote sensing |
| url | https://doi.org/10.1038/s41598-025-01476-3 |
| work_keys_str_mv | AT muhammadasif anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT mohammadabrar anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT faizanullah anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT abdusalam anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT farhanamin anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT isabeldelatorre anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT monicagraciavillar anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT helenagaray anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT gyusangchoi anovelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT muhammadasif novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT mohammadabrar novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT faizanullah novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT abdusalam novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT farhanamin novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT isabeldelatorre novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT monicagraciavillar novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT helenagaray novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing AT gyusangchoi novelhybriddeeplearningapproachforsuperresolutionandobjectsdetectioninremotesensing |