CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification
An important means of remote sensing (RS) imagery interpretation, RS scene classification technology, has recently achieved great success, especially based on deep learning. However, most of these methods are designed for noise-free images. The scene classification performance for noisy RS images, i...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9600873/ |
| _version_ | 1818683916511346688 |
|---|---|
| author | Bo Su Jun Liu Xin Su Bin Luo Qing Wang |
| author_facet | Bo Su Jun Liu Xin Su Bin Luo Qing Wang |
| author_sort | Bo Su |
| collection | DOAJ |
| description | An important means of remote sensing (RS) imagery interpretation, RS scene classification technology, has recently achieved great success, especially based on deep learning. However, most of these methods are designed for noise-free images. The scene classification performance for noisy RS images, i.e., synthetic aperture radar (SAR) images with speckle noise, is poor due to the sufficient effect of noise. An intuitive solution is denoising first and then classifying the image, which makes the whole process cumbersome. To address this problem, we design a new complete frequency channel attention network (CFCANet) that can handle noisy RS images directly without any filtering operation. CFCANet selects part of the low-frequency information to interact with the feature map adequately. For the original feature map, a corresponding 2-D discrete cosine transformation frequency component is assigned, from which the most significant eigenvalue of each channel is obtained by maximization. Compared with the frequency channel attention network (FcaNet), the proposed network has better antinoise ability as it exploits low frequency information of the images. The effectiveness of our method has been proved by experiments based on public datasets and some simulated datasets. Moreover, we build a new SAR scene classification dataset: WHU-SAR6. The comprehensive evaluation shows that the proposed method consistently outperforms several advanced methods, including ResNet, SENet, CBAM, EcaNet, and FcaNet. |
| first_indexed | 2024-12-17T10:42:20Z |
| format | Article |
| id | doaj.art-8f059e91f99948b581362a8e334dbc4b |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-12-17T10:42:20Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-8f059e91f99948b581362a8e334dbc4b2022-12-21T21:52:13ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352021-01-0114117501176310.1109/JSTARS.2021.31251079600873CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene ClassificationBo Su0https://orcid.org/0000-0001-5635-9117Jun Liu1https://orcid.org/0000-0002-8943-079XXin Su2https://orcid.org/0000-0003-0957-4628Bin Luo3https://orcid.org/0000-0002-3040-3500Qing Wang4School of Remote Sensing and Information Engineering, Wuhan University, Wuhan, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing, Wuhan University, Wuhan, ChinaSchool of Remote Sensing and Information Engineering, Wuhan University, Wuhan, ChinaState Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing, Wuhan University, Wuhan, ChinaAir Force Research Institute, Beijing, ChinaAn important means of remote sensing (RS) imagery interpretation, RS scene classification technology, has recently achieved great success, especially based on deep learning. However, most of these methods are designed for noise-free images. The scene classification performance for noisy RS images, i.e., synthetic aperture radar (SAR) images with speckle noise, is poor due to the sufficient effect of noise. An intuitive solution is denoising first and then classifying the image, which makes the whole process cumbersome. To address this problem, we design a new complete frequency channel attention network (CFCANet) that can handle noisy RS images directly without any filtering operation. CFCANet selects part of the low-frequency information to interact with the feature map adequately. For the original feature map, a corresponding 2-D discrete cosine transformation frequency component is assigned, from which the most significant eigenvalue of each channel is obtained by maximization. Compared with the frequency channel attention network (FcaNet), the proposed network has better antinoise ability as it exploits low frequency information of the images. The effectiveness of our method has been proved by experiments based on public datasets and some simulated datasets. Moreover, we build a new SAR scene classification dataset: WHU-SAR6. The comprehensive evaluation shows that the proposed method consistently outperforms several advanced methods, including ResNet, SENet, CBAM, EcaNet, and FcaNet.https://ieeexplore.ieee.org/document/9600873/Antinoisefrequency attention mechanismsynthetic aperture radar (SAR) image scene classification |
| spellingShingle | Bo Su Jun Liu Xin Su Bin Luo Qing Wang CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Antinoise frequency attention mechanism synthetic aperture radar (SAR) image scene classification |
| title | CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification |
| title_full | CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification |
| title_fullStr | CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification |
| title_full_unstemmed | CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification |
| title_short | CFCANet: A Complete Frequency Channel Attention Network for SAR Image Scene Classification |
| title_sort | cfcanet a complete frequency channel attention network for sar image scene classification |
| topic | Antinoise frequency attention mechanism synthetic aperture radar (SAR) image scene classification |
| url | https://ieeexplore.ieee.org/document/9600873/ |
| work_keys_str_mv | AT bosu cfcanetacompletefrequencychannelattentionnetworkforsarimagesceneclassification AT junliu cfcanetacompletefrequencychannelattentionnetworkforsarimagesceneclassification AT xinsu cfcanetacompletefrequencychannelattentionnetworkforsarimagesceneclassification AT binluo cfcanetacompletefrequencychannelattentionnetworkforsarimagesceneclassification AT qingwang cfcanetacompletefrequencychannelattentionnetworkforsarimagesceneclassification |