AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION
High-resolution images for remote sensing applications are often not affordable or accessible, especially when in need of a wide temporal span of recordings. Given the easy access to low-resolution (LR) images from satellites, many remote sensing works rely on this type of data. The problem is that...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2020-11-01
|
| Series: | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| Online Access: | https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLII-3-W12-2020/61/2020/isprs-archives-XLII-3-W12-2020-61-2020.pdf |
| _version_ | 1811292314963804160 |
|---|---|
| author | M. B. Pereira J. A. dos Santos |
| author_facet | M. B. Pereira J. A. dos Santos |
| author_sort | M. B. Pereira |
| collection | DOAJ |
| description | High-resolution images for remote sensing applications are often not affordable or accessible, especially when in need of a wide temporal span of recordings. Given the easy access to low-resolution (LR) images from satellites, many remote sensing works rely on this type of data. The problem is that LR images are not appropriate for semantic segmentation, due to the need for high-quality data for accurate pixel prediction for this task. In this paper, we propose an end-to-end framework that unites a super-resolution and a semantic segmentation module in order to produce accurate thematic maps from LR inputs. It allows the semantic segmentation network to conduct the reconstruction process, modifying the input image with helpful textures. We evaluate the framework with three remote sensing datasets. The results show that the framework is capable of achieving a semantic segmentation performance close to native high-resolution data, while also surpassing the performance of a network trained with LR inputs. |
| first_indexed | 2024-04-13T04:43:35Z |
| format | Article |
| id | doaj.art-a6409b04f81d4635a31978e517b22289 |
| institution | Directory Open Access Journal |
| issn | 1682-1750 2194-9034 |
| language | English |
| last_indexed | 2024-04-13T04:43:35Z |
| publishDate | 2020-11-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| spelling | doaj.art-a6409b04f81d4635a31978e517b222892022-12-22T03:01:55ZengCopernicus PublicationsThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences1682-17502194-90342020-11-01XLII-3-W12-2020616610.5194/isprs-archives-XLII-3-W12-2020-61-2020AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATIONM. B. Pereira0J. A. dos Santos1Dept. of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, BrazilDept. of Computer Science, Universidade Federal de Minas Gerais, Belo Horizonte, BrazilHigh-resolution images for remote sensing applications are often not affordable or accessible, especially when in need of a wide temporal span of recordings. Given the easy access to low-resolution (LR) images from satellites, many remote sensing works rely on this type of data. The problem is that LR images are not appropriate for semantic segmentation, due to the need for high-quality data for accurate pixel prediction for this task. In this paper, we propose an end-to-end framework that unites a super-resolution and a semantic segmentation module in order to produce accurate thematic maps from LR inputs. It allows the semantic segmentation network to conduct the reconstruction process, modifying the input image with helpful textures. We evaluate the framework with three remote sensing datasets. The results show that the framework is capable of achieving a semantic segmentation performance close to native high-resolution data, while also surpassing the performance of a network trained with LR inputs.https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLII-3-W12-2020/61/2020/isprs-archives-XLII-3-W12-2020-61-2020.pdf |
| spellingShingle | M. B. Pereira J. A. dos Santos AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences |
| title | AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION |
| title_full | AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION |
| title_fullStr | AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION |
| title_full_unstemmed | AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION |
| title_short | AN END-TO-END FRAMEWORK FOR LOW-RESOLUTION REMOTE SENSING SEMANTIC SEGMENTATION |
| title_sort | end to end framework for low resolution remote sensing semantic segmentation |
| url | https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLII-3-W12-2020/61/2020/isprs-archives-XLII-3-W12-2020-61-2020.pdf |
| work_keys_str_mv | AT mbpereira anendtoendframeworkforlowresolutionremotesensingsemanticsegmentation AT jadossantos anendtoendframeworkforlowresolutionremotesensingsemanticsegmentation AT mbpereira endtoendframeworkforlowresolutionremotesensingsemanticsegmentation AT jadossantos endtoendframeworkforlowresolutionremotesensingsemanticsegmentation |