BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images
Oil spill pollution at sea causes significant damage to marine ecosystems. Quad-polarimetric Synthetic Aperture Radar (SAR) has become an essential technology since it can provide polarization features for marine oil spill detection. Using deep learning models based on polarimetric features, oil spi...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-01-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/14/2/264 |
_version_ | 1797490753168998400 |
---|---|
author | Dawei Wang Jianhua Wan Shanwei Liu Yanlong Chen Muhammad Yasir Mingming Xu Peng Ren |
author_facet | Dawei Wang Jianhua Wan Shanwei Liu Yanlong Chen Muhammad Yasir Mingming Xu Peng Ren |
author_sort | Dawei Wang |
collection | DOAJ |
description | Oil spill pollution at sea causes significant damage to marine ecosystems. Quad-polarimetric Synthetic Aperture Radar (SAR) has become an essential technology since it can provide polarization features for marine oil spill detection. Using deep learning models based on polarimetric features, oil spill detection can be achieved. However, there is insufficient feature extraction due to model depth, small reception field lend due to loss of target information, and fixed hyperparameter for models. The effect of oil spill detection is still incomplete or misclassified. To solve the above problems, we propose an improved deep learning model named BO-DRNet. The model can obtain a more sufficiently and fuller feature by ResNet-18 as the backbone in encoder of DeepLabv3+, and Bayesian Optimization (BO) was used to optimize the model’s hyperparameters. Experiments were conducted based on ten prominent polarimetric features were extracted from three quad-polarimetric SAR images obtained by RADARSAT-2. Experimental results show that compared with other deep learning models, BO-DRNet performs best with a mean accuracy of 74.69% and a mean dice of 0.8551. This paper provides a valuable tool to manage upcoming disasters effectively. |
first_indexed | 2024-03-10T00:37:28Z |
format | Article |
id | doaj.art-9fdfcf6e2db74fa6a9c9e4466fcb406f |
institution | Directory Open Access Journal |
issn | 2072-4292 |
language | English |
last_indexed | 2024-03-10T00:37:28Z |
publishDate | 2022-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Remote Sensing |
spelling | doaj.art-9fdfcf6e2db74fa6a9c9e4466fcb406f2023-11-23T15:14:44ZengMDPI AGRemote Sensing2072-42922022-01-0114226410.3390/rs14020264BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR ImagesDawei Wang0Jianhua Wan1Shanwei Liu2Yanlong Chen3Muhammad Yasir4Mingming Xu5Peng Ren6College of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaCollege of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaCollege of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaNational Marine Environmental Monitoring Center, Dalian 116023, ChinaCollege of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaCollege of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaCollege of Oceanography and Space Informatics, China University of Petroleum, Qingdao 266580, ChinaOil spill pollution at sea causes significant damage to marine ecosystems. Quad-polarimetric Synthetic Aperture Radar (SAR) has become an essential technology since it can provide polarization features for marine oil spill detection. Using deep learning models based on polarimetric features, oil spill detection can be achieved. However, there is insufficient feature extraction due to model depth, small reception field lend due to loss of target information, and fixed hyperparameter for models. The effect of oil spill detection is still incomplete or misclassified. To solve the above problems, we propose an improved deep learning model named BO-DRNet. The model can obtain a more sufficiently and fuller feature by ResNet-18 as the backbone in encoder of DeepLabv3+, and Bayesian Optimization (BO) was used to optimize the model’s hyperparameters. Experiments were conducted based on ten prominent polarimetric features were extracted from three quad-polarimetric SAR images obtained by RADARSAT-2. Experimental results show that compared with other deep learning models, BO-DRNet performs best with a mean accuracy of 74.69% and a mean dice of 0.8551. This paper provides a valuable tool to manage upcoming disasters effectively.https://www.mdpi.com/2072-4292/14/2/264deep learning modeloil spill detectionpolarization featureSAR images |
spellingShingle | Dawei Wang Jianhua Wan Shanwei Liu Yanlong Chen Muhammad Yasir Mingming Xu Peng Ren BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images Remote Sensing deep learning model oil spill detection polarization feature SAR images |
title | BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images |
title_full | BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images |
title_fullStr | BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images |
title_full_unstemmed | BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images |
title_short | BO-DRNet: An Improved Deep Learning Model for Oil Spill Detection by Polarimetric Features from SAR Images |
title_sort | bo drnet an improved deep learning model for oil spill detection by polarimetric features from sar images |
topic | deep learning model oil spill detection polarization feature SAR images |
url | https://www.mdpi.com/2072-4292/14/2/264 |
work_keys_str_mv | AT daweiwang bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT jianhuawan bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT shanweiliu bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT yanlongchen bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT muhammadyasir bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT mingmingxu bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages AT pengren bodrnetanimproveddeeplearningmodelforoilspilldetectionbypolarimetricfeaturesfromsarimages |