DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection
Precise statistics on the spatial distribution of photovoltaics (PV) are essential for advancing the PV industry, and integrating remote sensing with artificial intelligence technologies offers a robust solution for accurate identification. Currently, numerous studies focus on the detection of singl...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/2/332 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832587608370184192 |
|---|---|
| author | Shaofu Lin Yang Yang Xiliang Liu Li Tian |
| author_facet | Shaofu Lin Yang Yang Xiliang Liu Li Tian |
| author_sort | Shaofu Lin |
| collection | DOAJ |
| description | Precise statistics on the spatial distribution of photovoltaics (PV) are essential for advancing the PV industry, and integrating remote sensing with artificial intelligence technologies offers a robust solution for accurate identification. Currently, numerous studies focus on the detection of single-type PV installations through aerial or satellite imagery. However, due to the variability in scale and shape of PV installations in complex environments, the detection results often fail to capture detailed information and struggle to scale for multi-scale PV systems. To tackle these challenges, a detection method known as Dynamic Spatial-Frequency Attention SwinNet (DSFA-SwinNet) for multi-scale PV areas is proposed. First, this study proposes the Dynamic Spatial-Frequency Attention (DSFA) mechanism, the Pyramid Attention Refinement (PAR) bottleneck structure, and optimizes the feature propagation method to achieve dynamic decoupling of the spatial and frequency domains in multi-scale representation learning. Secondly, a hybrid loss function has been developed with weights optimized employing the Bayesian Optimization algorithm to provide a strategic method for parameter tuning in similar research. Lastly, the fixed window size of Swin-Transformer is dynamically adjusted to enhance computational efficiency and maintain accuracy. The results on two PV datasets demonstrate that DSFA-SwinNet significantly enhances detection accuracy and scalability for multi-scale PV areas. |
| format | Article |
| id | doaj-art-f08311a2d339450ea747cd14a3e69535 |
| institution | Kabale University |
| issn | 2072-4292 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-f08311a2d339450ea747cd14a3e695352025-01-24T13:48:09ZengMDPI AGRemote Sensing2072-42922025-01-0117233210.3390/rs17020332DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas DetectionShaofu Lin0Yang Yang1Xiliang Liu2Li Tian3College of Computer Science, Beijing University of Technology, Chaoyang District, Beijing 100124, ChinaCollege of Computer Science, Beijing University of Technology, Chaoyang District, Beijing 100124, ChinaCollege of Computer Science, Beijing University of Technology, Chaoyang District, Beijing 100124, ChinaThe Institute of Geographic Sciences and Natural Resources Research, University of Chinese Academy of Sciences, Beijing 100101, ChinaPrecise statistics on the spatial distribution of photovoltaics (PV) are essential for advancing the PV industry, and integrating remote sensing with artificial intelligence technologies offers a robust solution for accurate identification. Currently, numerous studies focus on the detection of single-type PV installations through aerial or satellite imagery. However, due to the variability in scale and shape of PV installations in complex environments, the detection results often fail to capture detailed information and struggle to scale for multi-scale PV systems. To tackle these challenges, a detection method known as Dynamic Spatial-Frequency Attention SwinNet (DSFA-SwinNet) for multi-scale PV areas is proposed. First, this study proposes the Dynamic Spatial-Frequency Attention (DSFA) mechanism, the Pyramid Attention Refinement (PAR) bottleneck structure, and optimizes the feature propagation method to achieve dynamic decoupling of the spatial and frequency domains in multi-scale representation learning. Secondly, a hybrid loss function has been developed with weights optimized employing the Bayesian Optimization algorithm to provide a strategic method for parameter tuning in similar research. Lastly, the fixed window size of Swin-Transformer is dynamically adjusted to enhance computational efficiency and maintain accuracy. The results on two PV datasets demonstrate that DSFA-SwinNet significantly enhances detection accuracy and scalability for multi-scale PV areas.https://www.mdpi.com/2072-4292/17/2/332high-resolution imagesphotovoltaicswin-transformerdynamic spatial-frequency attention |
| spellingShingle | Shaofu Lin Yang Yang Xiliang Liu Li Tian DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection Remote Sensing high-resolution images photovoltaic swin-transformer dynamic spatial-frequency attention |
| title | DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection |
| title_full | DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection |
| title_fullStr | DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection |
| title_full_unstemmed | DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection |
| title_short | DSFA-SwinNet: A Multi-Scale Attention Fusion Network for Photovoltaic Areas Detection |
| title_sort | dsfa swinnet a multi scale attention fusion network for photovoltaic areas detection |
| topic | high-resolution images photovoltaic swin-transformer dynamic spatial-frequency attention |
| url | https://www.mdpi.com/2072-4292/17/2/332 |
| work_keys_str_mv | AT shaofulin dsfaswinnetamultiscaleattentionfusionnetworkforphotovoltaicareasdetection AT yangyang dsfaswinnetamultiscaleattentionfusionnetworkforphotovoltaicareasdetection AT xiliangliu dsfaswinnetamultiscaleattentionfusionnetworkforphotovoltaicareasdetection AT litian dsfaswinnetamultiscaleattentionfusionnetworkforphotovoltaicareasdetection |