HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline
Deep learning-based coastline detection algorithms have begun to outshine traditional statistical methods in recent years. However, they are usually trained only as single-purpose models to either segment land and water or delineate the coastline. In contrast to this, a human annotator will usually...
Main Authors: | , , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
arXiv
2021
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2103.01849 https://arxiv.org/abs/2103.01849 |
id |
ftdatacite:10.48550/arxiv.2103.01849 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2103.01849 2023-05-15T14:03:19+02:00 HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline Heidler, Konrad Mou, Lichao Baumhoer, Celia Dietz, Andreas Zhu, Xiao Xiang 2021 https://dx.doi.org/10.48550/arxiv.2103.01849 https://arxiv.org/abs/2103.01849 unknown arXiv https://dx.doi.org/10.1109/tgrs.2021.3064606 arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Computer Vision and Pattern Recognition cs.CV Image and Video Processing eess.IV FOS Computer and information sciences FOS Electrical engineering, electronic engineering, information engineering article-journal Article ScholarlyArticle Text 2021 ftdatacite https://doi.org/10.48550/arxiv.2103.01849 https://doi.org/10.1109/tgrs.2021.3064606 2022-03-10T14:24:20Z Deep learning-based coastline detection algorithms have begun to outshine traditional statistical methods in recent years. However, they are usually trained only as single-purpose models to either segment land and water or delineate the coastline. In contrast to this, a human annotator will usually keep a mental map of both segmentation and delineation when performing manual coastline detection. To take into account this task duality, we therefore devise a new model to unite these two approaches in a deep learning model. By taking inspiration from the main building blocks of a semantic segmentation framework (UNet) and an edge detection framework (HED), both tasks are combined in a natural way. Training is made efficient by employing deep supervision on side predictions at multiple resolutions. Finally, a hierarchical attention mechanism is introduced to adaptively merge these multiscale predictions into the final model output. The advantages of this approach over other traditional and deep learning-based methods for coastline detection are demonstrated on a dataset of Sentinel-1 imagery covering parts of the Antarctic coast, where coastline detection is notoriously difficult. An implementation of our method is available at \url{https://github.com/khdlr/HED-UNet}. : This work has been accepted by IEEE TGRS for publication. Copyright may be transferred without notice, after which this version may no longer be accessible Article in Journal/Newspaper Antarc* Antarctic DataCite Metadata Store (German National Library of Science and Technology) Antarctic The Antarctic |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Computer Vision and Pattern Recognition cs.CV Image and Video Processing eess.IV FOS Computer and information sciences FOS Electrical engineering, electronic engineering, information engineering |
spellingShingle |
Computer Vision and Pattern Recognition cs.CV Image and Video Processing eess.IV FOS Computer and information sciences FOS Electrical engineering, electronic engineering, information engineering Heidler, Konrad Mou, Lichao Baumhoer, Celia Dietz, Andreas Zhu, Xiao Xiang HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
topic_facet |
Computer Vision and Pattern Recognition cs.CV Image and Video Processing eess.IV FOS Computer and information sciences FOS Electrical engineering, electronic engineering, information engineering |
description |
Deep learning-based coastline detection algorithms have begun to outshine traditional statistical methods in recent years. However, they are usually trained only as single-purpose models to either segment land and water or delineate the coastline. In contrast to this, a human annotator will usually keep a mental map of both segmentation and delineation when performing manual coastline detection. To take into account this task duality, we therefore devise a new model to unite these two approaches in a deep learning model. By taking inspiration from the main building blocks of a semantic segmentation framework (UNet) and an edge detection framework (HED), both tasks are combined in a natural way. Training is made efficient by employing deep supervision on side predictions at multiple resolutions. Finally, a hierarchical attention mechanism is introduced to adaptively merge these multiscale predictions into the final model output. The advantages of this approach over other traditional and deep learning-based methods for coastline detection are demonstrated on a dataset of Sentinel-1 imagery covering parts of the Antarctic coast, where coastline detection is notoriously difficult. An implementation of our method is available at \url{https://github.com/khdlr/HED-UNet}. : This work has been accepted by IEEE TGRS for publication. Copyright may be transferred without notice, after which this version may no longer be accessible |
format |
Article in Journal/Newspaper |
author |
Heidler, Konrad Mou, Lichao Baumhoer, Celia Dietz, Andreas Zhu, Xiao Xiang |
author_facet |
Heidler, Konrad Mou, Lichao Baumhoer, Celia Dietz, Andreas Zhu, Xiao Xiang |
author_sort |
Heidler, Konrad |
title |
HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
title_short |
HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
title_full |
HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
title_fullStr |
HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
title_full_unstemmed |
HED-UNet: Combined Segmentation and Edge Detection for Monitoring the Antarctic Coastline |
title_sort |
hed-unet: combined segmentation and edge detection for monitoring the antarctic coastline |
publisher |
arXiv |
publishDate |
2021 |
url |
https://dx.doi.org/10.48550/arxiv.2103.01849 https://arxiv.org/abs/2103.01849 |
geographic |
Antarctic The Antarctic |
geographic_facet |
Antarctic The Antarctic |
genre |
Antarc* Antarctic |
genre_facet |
Antarc* Antarctic |
op_relation |
https://dx.doi.org/10.1109/tgrs.2021.3064606 |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2103.01849 https://doi.org/10.1109/tgrs.2021.3064606 |
_version_ |
1766273941903507456 |