Pretrained Detail Enhancement Framework for Remote Sensing Object Detection
Remote sensing object detection faces significant challenges due to the varying resolutions of images and the diverse shapes of objects. In this paper, we introduce an innovative approach to enhance object detection networks in remote sensing images by addressing these issues through a novel two-sta...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10829622/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841542566862913536 |
|---|---|
| author | Mo Zhou Yue Zhou Dawei Yang Kai Song |
| author_facet | Mo Zhou Yue Zhou Dawei Yang Kai Song |
| author_sort | Mo Zhou |
| collection | DOAJ |
| description | Remote sensing object detection faces significant challenges due to the varying resolutions of images and the diverse shapes of objects. In this paper, we introduce an innovative approach to enhance object detection networks in remote sensing images by addressing these issues through a novel two-stage training process. Our key contributions include the development of a combined low-resolution pretraining and high-resolution fine-tuning strategy, which significantly improves the model’s generalization capability across different resolutions. We have designed a Pretrained Detail Enhancement and Restoration (PDER) framework that enhances feature extraction from low-resolution images, thereby boosting the baseline network’s ability to handle blurred and intricate objects. Additionally, we propose an Energy-Efficient Shape Adaptive Convolution (EESA) that integrates a depthwise separable mechanism with shape-adaptive feature maps, enabling efficient capture of complex object edges without substantial computational overhead. Compared to state-of-the-art methods such as YOLOv7, our model achieves an absolute average precision improvement of 2.34% and 2.57% on the DIOR and xView datasets, respectively. Extensive experiments on multiple remote sensing datasets demonstrate our approach’s effectiveness, achieving significant improvements in detection accuracy and computational efficiency, thus highlighting its potential for practical applications. |
| format | Article |
| id | doaj-art-8f7a91a577694ef49d57d577c290f9bb |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-8f7a91a577694ef49d57d577c290f9bb2025-01-14T00:02:12ZengIEEEIEEE Access2169-35362025-01-01136362637710.1109/ACCESS.2025.352618010829622Pretrained Detail Enhancement Framework for Remote Sensing Object DetectionMo Zhou0https://orcid.org/0009-0002-8926-3722Yue Zhou1Dawei Yang2https://orcid.org/0000-0003-0717-9579Kai Song3https://orcid.org/0000-0002-8735-1609School Information Science and Engineering, Shenyang Ligong University, Shenyang, Liaoning, ChinaSchool Information Science and Engineering, Shenyang Ligong University, Shenyang, Liaoning, ChinaSchool Information Science and Engineering, Shenyang Ligong University, Shenyang, Liaoning, ChinaSchool Information Science and Engineering, Shenyang Ligong University, Shenyang, Liaoning, ChinaRemote sensing object detection faces significant challenges due to the varying resolutions of images and the diverse shapes of objects. In this paper, we introduce an innovative approach to enhance object detection networks in remote sensing images by addressing these issues through a novel two-stage training process. Our key contributions include the development of a combined low-resolution pretraining and high-resolution fine-tuning strategy, which significantly improves the model’s generalization capability across different resolutions. We have designed a Pretrained Detail Enhancement and Restoration (PDER) framework that enhances feature extraction from low-resolution images, thereby boosting the baseline network’s ability to handle blurred and intricate objects. Additionally, we propose an Energy-Efficient Shape Adaptive Convolution (EESA) that integrates a depthwise separable mechanism with shape-adaptive feature maps, enabling efficient capture of complex object edges without substantial computational overhead. Compared to state-of-the-art methods such as YOLOv7, our model achieves an absolute average precision improvement of 2.34% and 2.57% on the DIOR and xView datasets, respectively. Extensive experiments on multiple remote sensing datasets demonstrate our approach’s effectiveness, achieving significant improvements in detection accuracy and computational efficiency, thus highlighting its potential for practical applications.https://ieeexplore.ieee.org/document/10829622/Remote sensingobject detectiontransfer learningsuper-resolution |
| spellingShingle | Mo Zhou Yue Zhou Dawei Yang Kai Song Pretrained Detail Enhancement Framework for Remote Sensing Object Detection IEEE Access Remote sensing object detection transfer learning super-resolution |
| title | Pretrained Detail Enhancement Framework for Remote Sensing Object Detection |
| title_full | Pretrained Detail Enhancement Framework for Remote Sensing Object Detection |
| title_fullStr | Pretrained Detail Enhancement Framework for Remote Sensing Object Detection |
| title_full_unstemmed | Pretrained Detail Enhancement Framework for Remote Sensing Object Detection |
| title_short | Pretrained Detail Enhancement Framework for Remote Sensing Object Detection |
| title_sort | pretrained detail enhancement framework for remote sensing object detection |
| topic | Remote sensing object detection transfer learning super-resolution |
| url | https://ieeexplore.ieee.org/document/10829622/ |
| work_keys_str_mv | AT mozhou pretraineddetailenhancementframeworkforremotesensingobjectdetection AT yuezhou pretraineddetailenhancementframeworkforremotesensingobjectdetection AT daweiyang pretraineddetailenhancementframeworkforremotesensingobjectdetection AT kaisong pretraineddetailenhancementframeworkforremotesensingobjectdetection |