A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images

Oil spills have devastating impacts on ocean ecosystems, leading to significant economic losses. Consequently, it is crucial to develop effective techniques for the rapid detection of marine oil spills using satellite observations. Existing methods for detecting oil spills using synthetic aperture r...

Full description

Saved in:
Bibliographic Details
Main Authors: Xudong Huang, Biao Zhang, William Perrie
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10990149/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849325314713321472
author Xudong Huang
Biao Zhang
William Perrie
author_facet Xudong Huang
Biao Zhang
William Perrie
author_sort Xudong Huang
collection DOAJ
description Oil spills have devastating impacts on ocean ecosystems, leading to significant economic losses. Consequently, it is crucial to develop effective techniques for the rapid detection of marine oil spills using satellite observations. Existing methods for detecting oil spills using synthetic aperture radar (SAR) often face challenges related to algorithm complexity, limited data availability, and practical implementation. To address these issues, we propose a two-stage deep learning (DL) method to automatically localize and segment oil spills in Sentinel-1 SAR images. First, we utilize the faster R-CNN model to identify oil spill locations. Subsequently, subimages containing the oil spills are input into the UNet<sup>&#x002B;&#x002B;</sup> model to achieve pixel-wise segmentation. We have created a comprehensive dataset consisting of 3166 well-annotated oil spill samples derived from 1000 Sentinel-1 VV-polarized SAR images. This unique dataset will be used for training, validation, and testing the two DL models. Experimental results indicate that the precision and recall for oil spill localization are 87.46&#x0025; and 87.59&#x0025;, respectively, while for oil spill segmentation, they are 89.55&#x0025; and 89.37&#x0025;. The proposed method demonstrates optimal oil detection performance at wind speeds between 2 and 10 m&#x002F;s. Furthermore, the total runtime for oil spill localization and segmentation is less than 1 min for each full-resolution SAR image. These results suggest that the proposed approach has strong potential for operational oil spill monitoring using spaceborne SAR images.
format Article
id doaj-art-124e9146b83d410ca47dfc2b10add450
institution Kabale University
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-124e9146b83d410ca47dfc2b10add4502025-08-20T03:48:27ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-0118123151232710.1109/JSTARS.2025.356785910990149A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar ImagesXudong Huang0Biao Zhang1https://orcid.org/0000-0001-6569-1998William Perrie2State Key Laboratory of Climate System Prediction and Risk Management, Nanjing University of Information Science and Technology, Nanjing, ChinaSchool of Marine Sciences, Nanjing University of Information Science and Technology, Nanjing, ChinaFisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, NS, CanadaOil spills have devastating impacts on ocean ecosystems, leading to significant economic losses. Consequently, it is crucial to develop effective techniques for the rapid detection of marine oil spills using satellite observations. Existing methods for detecting oil spills using synthetic aperture radar (SAR) often face challenges related to algorithm complexity, limited data availability, and practical implementation. To address these issues, we propose a two-stage deep learning (DL) method to automatically localize and segment oil spills in Sentinel-1 SAR images. First, we utilize the faster R-CNN model to identify oil spill locations. Subsequently, subimages containing the oil spills are input into the UNet<sup>&#x002B;&#x002B;</sup> model to achieve pixel-wise segmentation. We have created a comprehensive dataset consisting of 3166 well-annotated oil spill samples derived from 1000 Sentinel-1 VV-polarized SAR images. This unique dataset will be used for training, validation, and testing the two DL models. Experimental results indicate that the precision and recall for oil spill localization are 87.46&#x0025; and 87.59&#x0025;, respectively, while for oil spill segmentation, they are 89.55&#x0025; and 89.37&#x0025;. The proposed method demonstrates optimal oil detection performance at wind speeds between 2 and 10 m&#x002F;s. Furthermore, the total runtime for oil spill localization and segmentation is less than 1 min for each full-resolution SAR image. These results suggest that the proposed approach has strong potential for operational oil spill monitoring using spaceborne SAR images.https://ieeexplore.ieee.org/document/10990149/Deep learning (DL)faster R-CNNoil spillsynthetic aperture radar (SAR)UNet<sup xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">++</sup>
spellingShingle Xudong Huang
Biao Zhang
William Perrie
A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Deep learning (DL)
faster R-CNN
oil spill
synthetic aperture radar (SAR)
UNet<sup xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">++</sup>
title A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
title_full A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
title_fullStr A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
title_full_unstemmed A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
title_short A Two-Stage Deep Learning Method for Marine Oil Spill Localization and Segmentation From Synthetic Aperture Radar Images
title_sort two stage deep learning method for marine oil spill localization and segmentation from synthetic aperture radar images
topic Deep learning (DL)
faster R-CNN
oil spill
synthetic aperture radar (SAR)
UNet<sup xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">++</sup>
url https://ieeexplore.ieee.org/document/10990149/
work_keys_str_mv AT xudonghuang atwostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages
AT biaozhang atwostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages
AT williamperrie atwostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages
AT xudonghuang twostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages
AT biaozhang twostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages
AT williamperrie twostagedeeplearningmethodformarineoilspilllocalizationandsegmentationfromsyntheticapertureradarimages