Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions
AbstractBuilding an oil spill segmentation model is very challenging because of the limited available information on oil spill accidents. Therefore, this paper proposes a custom data generator based on Segmentation Network (Seg-Net) model implemented in Conditional Generative Adversarial Network (CG...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Taylor & Francis Group
2023-12-01
|
Series: | Geomatics, Natural Hazards & Risk |
Subjects: | |
Online Access: | https://www.tandfonline.com/doi/10.1080/19475705.2022.2155998 |
_version_ | 1797389240594595840 |
---|---|
author | Samira Ahmed Tamer ElGharbawi Mahmoud Salah Mahmoud El-Mewafi |
author_facet | Samira Ahmed Tamer ElGharbawi Mahmoud Salah Mahmoud El-Mewafi |
author_sort | Samira Ahmed |
collection | DOAJ |
description | AbstractBuilding an oil spill segmentation model is very challenging because of the limited available information on oil spill accidents. Therefore, this paper proposes a custom data generator based on Segmentation Network (Seg-Net) model implemented in Conditional Generative Adversarial Network (CGAN). The proposed model is trained for oil spill segmentation using 50 Sentinal-1 Synthetic Aperture Radar (SAR) images. The proposed model employs a modified Seg-Net as a generator to produce high-quality oil spills’ images and a Patch-GAN as discriminator. This architecture results in a significant improvement of the final oil segmentation results, in comparison with Seg-Net model, while using relatively small training dataset. For performance assessment, the paper presents the oil spills segmentation results of four suggested models using Sentinel-1 SAR images. The presented models are U-Net, Seg-Net, CGAN, and a Seg-Net-based CGAN the performance assessment reveals that the proposed model produces oil spill segmentation images with an average accuracy of 99.04%, Intersection over Union (IoU) index of 96.59%, and a precision of 85.24%. In addition, the training time required for the proposed model is 3 h 20 min per 50 epochs, while it is nearly 10 h and 55 min for training a CGAN model. |
first_indexed | 2024-03-08T22:52:59Z |
format | Article |
id | doaj.art-e65c88ea075949af9a51f0051a083cd3 |
institution | Directory Open Access Journal |
issn | 1947-5705 1947-5713 |
language | English |
last_indexed | 2024-03-08T22:52:59Z |
publishDate | 2023-12-01 |
publisher | Taylor & Francis Group |
record_format | Article |
series | Geomatics, Natural Hazards & Risk |
spelling | doaj.art-e65c88ea075949af9a51f0051a083cd32023-12-16T08:49:46ZengTaylor & Francis GroupGeomatics, Natural Hazards & Risk1947-57051947-57132023-12-01141769410.1080/19475705.2022.2155998Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regionsSamira Ahmed0Tamer ElGharbawi1Mahmoud Salah2Mahmoud El-Mewafi3Public Works Engineering Department, Mansoura University, Mansoura, EgyptFaculty of Engineering, Suez Canal University, Ismailia, EgyptDepartment of Surveying Engineering, Faculty of Engineering Shoubra, Benha University, Banha, EgyptPublic Works Engineering Department, Mansoura University, Mansoura, EgyptAbstractBuilding an oil spill segmentation model is very challenging because of the limited available information on oil spill accidents. Therefore, this paper proposes a custom data generator based on Segmentation Network (Seg-Net) model implemented in Conditional Generative Adversarial Network (CGAN). The proposed model is trained for oil spill segmentation using 50 Sentinal-1 Synthetic Aperture Radar (SAR) images. The proposed model employs a modified Seg-Net as a generator to produce high-quality oil spills’ images and a Patch-GAN as discriminator. This architecture results in a significant improvement of the final oil segmentation results, in comparison with Seg-Net model, while using relatively small training dataset. For performance assessment, the paper presents the oil spills segmentation results of four suggested models using Sentinel-1 SAR images. The presented models are U-Net, Seg-Net, CGAN, and a Seg-Net-based CGAN the performance assessment reveals that the proposed model produces oil spill segmentation images with an average accuracy of 99.04%, Intersection over Union (IoU) index of 96.59%, and a precision of 85.24%. In addition, the training time required for the proposed model is 3 h 20 min per 50 epochs, while it is nearly 10 h and 55 min for training a CGAN model.https://www.tandfonline.com/doi/10.1080/19475705.2022.2155998Oil spillsSARdeep learningsegmentationgenerative adversarial network |
spellingShingle | Samira Ahmed Tamer ElGharbawi Mahmoud Salah Mahmoud El-Mewafi Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions Geomatics, Natural Hazards & Risk Oil spills SAR deep learning segmentation generative adversarial network |
title | Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions |
title_full | Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions |
title_fullStr | Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions |
title_full_unstemmed | Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions |
title_short | Deep neural network for oil spill detection using Sentinel-1 data: application to Egyptian coastal regions |
title_sort | deep neural network for oil spill detection using sentinel 1 data application to egyptian coastal regions |
topic | Oil spills SAR deep learning segmentation generative adversarial network |
url | https://www.tandfonline.com/doi/10.1080/19475705.2022.2155998 |
work_keys_str_mv | AT samiraahmed deepneuralnetworkforoilspilldetectionusingsentinel1dataapplicationtoegyptiancoastalregions AT tamerelgharbawi deepneuralnetworkforoilspilldetectionusingsentinel1dataapplicationtoegyptiancoastalregions AT mahmoudsalah deepneuralnetworkforoilspilldetectionusingsentinel1dataapplicationtoegyptiancoastalregions AT mahmoudelmewafi deepneuralnetworkforoilspilldetectionusingsentinel1dataapplicationtoegyptiancoastalregions |