dc.contributor.author |
Mdakane, Lizwe W
|
|
dc.contributor.author |
Kleynhans, W
|
|
dc.date.accessioned |
2021-01-17T15:51:51Z |
|
dc.date.available |
2021-01-17T15:51:51Z |
|
dc.date.issued |
2020-10 |
|
dc.identifier.citation |
•Mdakane, L.W. & Kleynhans, W. 2020. Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data. IEEE Geoscience and Remote Sensing Letters, pp. 12 |
en_US |
dc.identifier.issn |
1545-598X |
|
dc.identifier.issn |
1558-0571 |
|
dc.identifier.uri |
https://ieeexplore.ieee.org/document/9212584
|
|
dc.identifier.uri |
DOI: 10.1109/LGRS.2020.3025641
|
|
dc.identifier.uri |
http://hdl.handle.net/10204/11716
|
|
dc.description |
Copyright: 2020 IEEE. This is a pre-print version. The definitive version of the work is published in Mdakane, L.W. & Kleynhans, W. 2020. Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data. IEEE Geoscience and Remote Sensing Letters, pp. 12 |
en_US |
dc.description.abstract |
Oil spills are often caused by vessels when dumping oily bilge wastewater at sea (also referred to as bilge dumping). In an synthetic aperture radar (SAR) image, oil spills dampen the radar energy return and appear as linearly shaped dark regions. However, naturally occurring phenomena (e.g., natural seepage) known as oil spill look-alikes can also dampen energy return and occur more often compared to a real oil spill. The primary goal of the study is to develop a monitoring system dedicated to automatically detect oil spill events caused by ships (bilge dumping) in African Oceans. To achieve this goal, the knowledge of features that has a high probability of separating oil spills from look-alikes is of great importance. The study aimed to accomplish three things, 1) to improve the lack of oil spill studies in Africa; 2) to determine the critical features that yield the highest discrimination accuracy of oil spills caused by moving vessels from look-alikes; and 3) to use these features to automatically detect and classify oil spill events. The study investigated the most common features used in literature for discriminating oil spills from look-alikes from SAR imagery. Multiple feature selection methods and the gradient boosting decision tree (GBT) classifier were used to select, classify, and determine the significant features for discriminating oil spills caused by moving vessels. The results showed that while some features vary, there are features that consistently have high and low significance across all methods. |
en_US |
dc.language.iso |
en |
en_US |
dc.publisher |
IEEE |
en_US |
dc.relation.ispartofseries |
Workflow;23878 |
|
dc.subject |
Oils |
en_US |
dc.subject |
Feature extraction |
en_US |
dc.subject |
Synthetic aperture radar |
en_US |
dc.subject |
SAR |
en_US |
dc.subject |
Image segmentation |
en_US |
dc.subject |
Oceans |
en_US |
dc.subject |
Bilge waste dumping |
en_US |
dc.subject |
Object classification |
en_US |
dc.subject |
Oil spills |
en_US |
dc.title |
Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data |
en_US |
dc.type |
Article |
en_US |
dc.identifier.apacitation |
Mdakane, L. W., & Kleynhans, W. (2020). Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data. http://hdl.handle.net/10204/11716 |
en_ZA |
dc.identifier.chicagocitation |
Mdakane, Lizwe W, and W Kleynhans "Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data." (2020) http://hdl.handle.net/10204/11716 |
en_ZA |
dc.identifier.vancouvercitation |
Mdakane LW, Kleynhans W. Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data. 2020; http://hdl.handle.net/10204/11716. |
en_ZA |
dc.identifier.ris |
TY - Article
AU - Mdakane, Lizwe W
AU - Kleynhans, W
AB - Oil spills are often caused by vessels when dumping oily bilge wastewater at sea (also referred to as bilge dumping). In an synthetic aperture radar (SAR) image, oil spills dampen the radar energy return and appear as linearly shaped dark regions. However, naturally occurring phenomena (e.g., natural seepage) known as oil spill look-alikes can also dampen energy return and occur more often compared to a real oil spill. The primary goal of the study is to develop a monitoring system dedicated to automatically detect oil spill events caused by ships (bilge dumping) in African Oceans. To achieve this goal, the knowledge of features that has a high probability of separating oil spills from look-alikes is of great importance. The study aimed to accomplish three things, 1) to improve the lack of oil spill studies in Africa; 2) to determine the critical features that yield the highest discrimination accuracy of oil spills caused by moving vessels from look-alikes; and 3) to use these features to automatically detect and classify oil spill events. The study investigated the most common features used in literature for discriminating oil spills from look-alikes from SAR imagery. Multiple feature selection methods and the gradient boosting decision tree (GBT) classifier were used to select, classify, and determine the significant features for discriminating oil spills caused by moving vessels. The results showed that while some features vary, there are features that consistently have high and low significance across all methods.
DA - 2020-10
DB - ResearchSpace
DP - CSIR
KW - Oils
KW - Feature extraction
KW - Synthetic aperture radar
KW - SAR
KW - Image segmentation
KW - Oceans
KW - Bilge waste dumping
KW - Object classification
KW - Oil spills
LK - https://researchspace.csir.co.za
PY - 2020
SM - 1545-598X
SM - 1558-0571
T1 - Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data
TI - Feature selection and classification of oil spill from vessels using Sentinel-1 wide–swath synthetic aperture radar data
UR - http://hdl.handle.net/10204/11716
ER -
|
en_ZA |