CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks
We present a method for cloud removal from satellite images using axial transformer networks. The method considers a set of multi-temporal images in a given region of interest, together with the corresponding cloud masks, and produces a cloud-free image for a specific day of the year. We propose the...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/17/1/86 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841549000327561216 |
|---|---|
| author | Dionysis Christopoulos Valsamis Ntouskos Konstantinos Karantzalos |
| author_facet | Dionysis Christopoulos Valsamis Ntouskos Konstantinos Karantzalos |
| author_sort | Dionysis Christopoulos |
| collection | DOAJ |
| description | We present a method for cloud removal from satellite images using axial transformer networks. The method considers a set of multi-temporal images in a given region of interest, together with the corresponding cloud masks, and produces a cloud-free image for a specific day of the year. We propose the combination of an encoder-decoder model employing axial attention layers for the estimation of the low-resolution cloud-free image, together with a fully parallel upsampler that reconstructs the image at full resolution. The method is compared with various baselines and state-of-the-art methods on Sentinel-2 datasets of different coverage, showing significant improvements across multiple standard metrics used for image quality assessment. |
| format | Article |
| id | doaj-art-7b615e6704284103ace7f0c7160fc7f2 |
| institution | Kabale University |
| issn | 2072-4292 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj-art-7b615e6704284103ace7f0c7160fc7f22025-01-10T13:20:11ZengMDPI AGRemote Sensing2072-42922024-12-011718610.3390/rs17010086CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer NetworksDionysis Christopoulos0Valsamis Ntouskos1Konstantinos Karantzalos2Remote Sensing Lab, National Technical University of Athens, 157 72 Athens, GreeceRemote Sensing Lab, National Technical University of Athens, 157 72 Athens, GreeceRemote Sensing Lab, National Technical University of Athens, 157 72 Athens, GreeceWe present a method for cloud removal from satellite images using axial transformer networks. The method considers a set of multi-temporal images in a given region of interest, together with the corresponding cloud masks, and produces a cloud-free image for a specific day of the year. We propose the combination of an encoder-decoder model employing axial attention layers for the estimation of the low-resolution cloud-free image, together with a fully parallel upsampler that reconstructs the image at full resolution. The method is compared with various baselines and state-of-the-art methods on Sentinel-2 datasets of different coverage, showing significant improvements across multiple standard metrics used for image quality assessment.https://www.mdpi.com/2072-4292/17/1/86autoregressive modelscloud removalimage reconstructionmulti-temporaloptical imagingSentinel-2 |
| spellingShingle | Dionysis Christopoulos Valsamis Ntouskos Konstantinos Karantzalos CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks Remote Sensing autoregressive models cloud removal image reconstruction multi-temporal optical imaging Sentinel-2 |
| title | CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks |
| title_full | CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks |
| title_fullStr | CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks |
| title_full_unstemmed | CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks |
| title_short | CloudTran++: Improved Cloud Removal from Multi-Temporal Satellite Images Using Axial Transformer Networks |
| title_sort | cloudtran improved cloud removal from multi temporal satellite images using axial transformer networks |
| topic | autoregressive models cloud removal image reconstruction multi-temporal optical imaging Sentinel-2 |
| url | https://www.mdpi.com/2072-4292/17/1/86 |
| work_keys_str_mv | AT dionysischristopoulos cloudtranimprovedcloudremovalfrommultitemporalsatelliteimagesusingaxialtransformernetworks AT valsamisntouskos cloudtranimprovedcloudremovalfrommultitemporalsatelliteimagesusingaxialtransformernetworks AT konstantinoskarantzalos cloudtranimprovedcloudremovalfrommultitemporalsatelliteimagesusingaxialtransformernetworks |