Oil Spill Detection from PlanetScope Satellite Image: Application to Oil Spill Accident near Ras Al Zour Area, Kuwait in August 2017

Park, Sung-Hwan; Jung, Hyung-Sup; Lee, Moung-Jin; Lee, Won-Jin; Choi, Myoung-Jin

doi:10.2112/si90-031.1

Cited by 23 publications

(12 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent deep-learning-based methods [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] that used CNN structures for both automated feature extraction, as well as classification of SAR images have relied on the use of patches to reduce the background concentration in the tested images. Pre-trained models, such as ResNet 101, VGG-16, and GAN networks as in [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] or multi-level CNN networks as in [9], were introduced to classify patches with modest performance (i.e., precision and Dice score). In [23], authors obtained improved results via introducing the use of VGG-16 and dark batch generation algorithm on spaceborne SAR images.…”

Section: Discussionmentioning

confidence: 99%

“…The proposed approach achieved an accuracy above 85%. Park et al introduced the use of artificial neural networks to detect oil leaks in optical PlanetScope satellite images acquired close to Ras Al Zour town in Kuwait [21]. Sun glint effects and dust were subtracted from the images and then provided to an artificial neural network to classify the target pixels into three types of oil leaks and sea surfaces with an overall accuracy of 82%.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Deep-Learning Framework for the Detection of Oil Spills from SAR Data

Shaban

Salim

Khalifeh

et al. 2021

Sensors

View full text Add to dashboard Cite

Oil leaks onto water surfaces from big tankers, ships, and pipeline cracks cause considerable damage and harm to the marine environment. Synthetic Aperture Radar (SAR) images provide an approximate representation for target scenes, including sea and land surfaces, ships, oil spills, and look-alikes. Detection and segmentation of oil spills from SAR images are crucial to aid in leak cleanups and protecting the environment. This paper introduces a two-stage deep-learning framework for the identification of oil spill occurrences based on a highly unbalanced dataset. The first stage classifies patches based on the percentage of oil spill pixels using a novel 23-layer Convolutional Neural Network. In contrast, the second stage performs semantic segmentation using a five-stage U-Net structure. The generalized Dice loss is minimized to account for the reduced oil spill representation in the patches. The results of this study are very promising and provide a comparable improved precision and Dice score compared to related work.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

A Deep-Learning Framework for the Detection of Oil Spills from SAR Data

Shaban

Salim

Khalifeh

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…However, the processing time necessary to compute these vast amounts of data must be reduced to generate information quickly in an emergency situation. Park et al (2019), for example, applied an artificial neural network (ANN) for the detection of oil slick areas in Kuwait in August 2017, using a methodology to reduce the processing time. Sunlight wave effects were removed using a median filter previous to the ANN classification, which provided an overall accuracy of 82.01% with a Kappa coefficient of 72.42.…”

Section: Nanosatellite Applications For Earth Remote Sensingmentioning

confidence: 99%

Nanosatellites applied to optical Earth observation: a review

2020

View full text Add to dashboard Cite

Nanosatellites and CubeSats were first developed for educational purposes. However, their low cost and short development cycle made nanosatellite constellations an affordable option for observing the Earth by remote sensing, increasing the frequency of high-resolution imagery, which is fundamental for studying and monitoring dynamic processes. In this sense, although still incipient, nanosatellite applications and proposed Earth observation missions are steadily growing in number and scientific fields. There are several initiatives from universities, space agencies and private companies to launch new nanosatellite missions. These initiatives are actively investigating new technologies to improve image quality and studying ways to increase acquisition frequency through the launch of larger constellations. So far, the private sector is leading the development of new missions, with proposals ranging from 12 to more than one thousand nanosatellite constellations. Furthermore, new nanosatellite missions have been proposed to tackle specific applications, such as natural disasters, or to test improvements on nanosatellite spatial, temporal and radiometric resolution. The unprecedented combination of high spatial and temporal resolution from nanosatellite constellations associated with improvement efforts in sensor quality is promising and may represent a trend to replace the era of large satellites for smaller and cheaper nanosatellites. This article first reports on the development and new nanosatellite missions of space agencies, universities and private companies. Then a systematic review of published articles using the most successful private constellation (PlanetScope and Doves) is presented and the principal papers are discussed.

show abstract

“…And the environmental monitoring studies were also conducted from the OISST, ARGO, MODIS, Landsat, and TerraSAR-X images (Baek and Moon, 2019;Chen et al, 2019a;Eom et al, 2019;Hong et al, 2019;Jeong et al, 2019;Jung et al, 2019;Lee et al, 2019a;Li et al, 2019;Ma et al, 2019;Mu et al, 2019;Sun et al, 2019;Tong et al, 2019;Qing, Hao, and Bao, 2019;Ren et al, 2019b;Xiao, Zhang, and Qin, 2019;Zhang et al, 2019aZhang et al, , 2019b The research topics of the oil spill, typhoon, flood, and nuclear radiation emergent have been carried out by using optical and SAR images (Bing et al, 2019;Jin et al, 2019;Kim and Moon, 2019;Park et al, 2019b;Syifa et al, 2019;Yang et al, 2019). Moreover, the specific topics related to marine spatial planning have been studied (Achmad et al, 2019;Bae et al, 2019;Chu et al, 2019;Chun and Lee, 2019;Jang et al, 2019;Kim et al, 2019b;Kim, Baek, and Hwang, 2019;Ko and Lee, 2019;Koo et al, 2019;Lee et al, 2019bLee et al, , 2019cLee et al, , 2019dOh et al, 2019aOh et al, , 2019bPark, 2019;Park et al 2019c;Ren et al, ...…”

Section: Previous Special Issue Related To Geospatial Research Of Coamentioning

confidence: 99%