Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention
Object detection is an important task of remote sensing applications. In recent years, with the development of deep convolutional neural networks, object detection in remote sensing images has made great improvements. However, the large variation of object scales and complex scenarios will seriously...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-01-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/2/427 |
| _version_ | 1797490598747308032 |
|---|---|
| author | Jiahang Liu Donghao Yang Fei Hu |
| author_facet | Jiahang Liu Donghao Yang Fei Hu |
| author_sort | Jiahang Liu |
| collection | DOAJ |
| description | Object detection is an important task of remote sensing applications. In recent years, with the development of deep convolutional neural networks, object detection in remote sensing images has made great improvements. However, the large variation of object scales and complex scenarios will seriously affect the performance of the detectors. To solve these problems, a novel object detection algorithm based on multi-receptive-field features and relation-connected attention is proposed for remote sensing images to achieve more accurate detection results. Specifically, we propose a multi-receptive-field feature extraction module with dilated convolution to aggregate the context information of different receptive fields. This achieves a strong capability of feature representation, which can effectively adapt to the scale changes of objects, either due to various object scales or different resolutions. Then, a relation-connected attention module based on relation modeling is constructed to automatically select and refine the features, which combines global and local attention to make the features more discriminative and can effectively improve the robustness of the detector. We designed these two modules as plug-and-play blocks and integrated them into the framework of Faster R-CNN to verify our method. The experimental results on NWPU VHR-10 and HRSC2016 datasets demonstrate that these two modules can effectively improve the performance of basic deep CNNs, and the proposed method can achieve better results of multiscale object detection in complex backgrounds. |
| first_indexed | 2024-03-10T00:35:17Z |
| format | Article |
| id | doaj.art-561d7704b47e49b7892208ec6622b140 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T00:35:17Z |
| publishDate | 2022-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-561d7704b47e49b7892208ec6622b1402023-11-23T15:17:38ZengMDPI AGRemote Sensing2072-42922022-01-0114242710.3390/rs14020427Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected AttentionJiahang Liu0Donghao Yang1Fei Hu2College of Astronautics, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Astronautics, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Astronautics, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaObject detection is an important task of remote sensing applications. In recent years, with the development of deep convolutional neural networks, object detection in remote sensing images has made great improvements. However, the large variation of object scales and complex scenarios will seriously affect the performance of the detectors. To solve these problems, a novel object detection algorithm based on multi-receptive-field features and relation-connected attention is proposed for remote sensing images to achieve more accurate detection results. Specifically, we propose a multi-receptive-field feature extraction module with dilated convolution to aggregate the context information of different receptive fields. This achieves a strong capability of feature representation, which can effectively adapt to the scale changes of objects, either due to various object scales or different resolutions. Then, a relation-connected attention module based on relation modeling is constructed to automatically select and refine the features, which combines global and local attention to make the features more discriminative and can effectively improve the robustness of the detector. We designed these two modules as plug-and-play blocks and integrated them into the framework of Faster R-CNN to verify our method. The experimental results on NWPU VHR-10 and HRSC2016 datasets demonstrate that these two modules can effectively improve the performance of basic deep CNNs, and the proposed method can achieve better results of multiscale object detection in complex backgrounds.https://www.mdpi.com/2072-4292/14/2/427convolutional neural networks (CNNs)multi-receptive-field feature extractionmultiscale object detectionrelation-connected attentionremote sensing images |
| spellingShingle | Jiahang Liu Donghao Yang Fei Hu Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention Remote Sensing convolutional neural networks (CNNs) multi-receptive-field feature extraction multiscale object detection relation-connected attention remote sensing images |
| title | Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention |
| title_full | Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention |
| title_fullStr | Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention |
| title_full_unstemmed | Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention |
| title_short | Multiscale Object Detection in Remote Sensing Images Combined with Multi-Receptive-Field Features and Relation-Connected Attention |
| title_sort | multiscale object detection in remote sensing images combined with multi receptive field features and relation connected attention |
| topic | convolutional neural networks (CNNs) multi-receptive-field feature extraction multiscale object detection relation-connected attention remote sensing images |
| url | https://www.mdpi.com/2072-4292/14/2/427 |
| work_keys_str_mv | AT jiahangliu multiscaleobjectdetectioninremotesensingimagescombinedwithmultireceptivefieldfeaturesandrelationconnectedattention AT donghaoyang multiscaleobjectdetectioninremotesensingimagescombinedwithmultireceptivefieldfeaturesandrelationconnectedattention AT feihu multiscaleobjectdetectioninremotesensingimagescombinedwithmultireceptivefieldfeaturesandrelationconnectedattention |