Oil spills detection from SAR Earth observations based on a hybrid CNN transformer networks.

Oil spills are the main threats to marine and coastal environments. Due to the increase in the marine transportation and shipping industry, oil spills have increased in recent years. Moreover, the rapid spread of oil spills in open waters seriously affects the fragile marine ecosystem and creates environmental concerns. Effective monitoring, quick identification, and estimation of the volume of oil spills are the first and most crucial steps for a successful cleanup operation and crisis management. Remote Sensing observations, especially from Synthetic Aperture Radar (SAR) sensors, are a very suitable choice for this purpose due to their ability to collect data regardless of the weather and illumination conditions and over far and large areas of the Earth. Owing to the relatively complex nature of SAR observations, machine learning (ML) based algorithms play an important role in accurately detecting and monitoring oil spills and can significantly help experts in faster and more accurate detection. This paper uses SAR images from ESA's Copernicus Sentinel-1 satellite to detect and locate oil spills in open waters under different environmental conditions. To this end, a deep learning framework has been presented to identify oil spills automatically. The SAR images were segmented into two classes, the oil slick and the background, using convolutional neural networks (CNN) and vision transformers (ViT). Various scenarios for the proposed architecture were designed by placing ViT networks in different parts of the CNN backbone. An extensive dataset of oil spill events in various regions across the globe was used to train and assess the performance of the proposed framework. After the detection performance assessments, the F1-score values for the standard DeepLabV3+, FC-DenseNet, and U-Net networks were 75.08 %, 73.94 %, and 60.85, respectively. In the combined networks models (combination of CNN and ViT), the best F1-score results were obtained as 78.48 %. Our results showed that these hybrid models could improve detection accuracy and have a high ability to distinguish oil spill borders even in noisy images. Evaluation metrics are increased in all the combined networks compared to the original CNN networks.

Dehghani-Dehcheshmeh S ，Akhoondzadeh M ，Homayouni S 《-》

Application of C-band sentinel-1A SAR data as proxies for detecting oil spills of Chennai, East Coast of India.

This paper presents the utilization of Synthetic Aperture Radar (SAR) data for monitoring and detection of oil spills. In this work, a case study of an oil spill has been investigated using C-band Sentinel-1A SAR data to detect the oil spill that occurred on 28 January 2017, near Ennore port, Chennai, India. Oil spill damages marine ecosystems causing serious environmental effects. Quite often, oil spills on the sea/ocean surface are seen nowadays, mainly in major shipping routes. They are caused due to tanker collisions, illegal discharge from the ships, etc. An oil spill can be monitored and detected using various platforms such as vessel-based, airborne-based and satellite-based. Vessel based and airborne methods are expensive with less area coverage. This process also consumes more time. For ocean applications such as oil spill and Ship detection, optical sensors cannot image during bad weather. As SAR is an active sensor, weather independent, and has cloud penetrating capability, the images can be acquired during the day as well as at night. Radar Remote Sensing (RRS) has rapidly gained popularity for monitoring and detection of oil spills and ships for more than a decade. With the availability of the satellite images, detection of oil spill has improved due to its wide coverage and less revisit time. The present paper gives an overview of the methodologies used to detect oil spills on the SAR images using dual-pol Sentinel-1A Level 1 SLC data. This work clearly demonstrates the preprocessing steps of the Sentinel 1A data for oil spill detection. The oil spill was only visible in the VV channel, therefore, for ocean application VV channel image is preferred. SEASAT was the first space-borne SAR mission launched in 1978 by NASA to observe sea surface. The preprocessing was carried out at the European Space Agency (ESA), the Sentinel Application Platform (SNAP) toolbox and Envi 5.1 toolbox. Based on the Sigma naught values, oil spill can be discriminated with the ocean surface. The results obtained with the VV channel are satisfactory and one could map out the oil spill very well. Supervised classifiers SVM and NN were applied on the boxcar filtered 3 × 3 VV channel image to delineate the oil spill. The result of oil spill detection mapping is validated with Supervised SVM and Neural Network classifiers. The results show there is a good agreement between oil spill mapping and classified image using SVM and NN classified images. The Overall Accuracy (OA) obtained using SVM classifier is 98.13% with kappa coefficient as 0.95 and using NN classifier is 98.11% with kappa coefficients 0.95. This technique is considered to be a potential proxy for the detection and monitoring of Oil spills on water bodies. Application of SAR data for oil spill detection is considered to be first of its kind from Indian coasts. This study aims to detect the oil spill occurred due to collision of two LPG tankers with Sentinel-1A SLC data in Chennai coast area.

Dasari K ，Anjaneyulu L ，Nadimikeri J 《-》

A novel deep learning method for marine oil spill detection from satellite synthetic aperture radar imagery.

Oil spill discharges from operational maritime activities like ships, oil rigs and other structures, leaking pipelines, as well as natural hydrocarbon seepage pose serious threats to marine ecosystems and fisheries. Satellite synthetic aperture radar (SAR) is a unique microwave instrument for marine oil spill monitoring, as it is not dependent on weather or sunlight conditions. Existing SAR oil spill detection approaches are limited by algorithm complexity, imbalanced data sets, uncertainties in selecting optimal features, and relatively slow detection speed. To overcome these restrictions, a fast and effective SAR oil spill detection method is presented, based a novel deep learning model, named the Faster Region-based Convolutional Neural Network (Faster R-CNN). This approach is capable of achieving fast end-to-end oil spill detection with reasonable accuracy. A large data set consisting of 15,774 labeled oil spill samples derived from 1786C-band Sentinel-1 and RADARSAT-2 vertical polarization SAR images is used to train, validate and test the Faster R-CNN model. Our experimental results show that the proposed method exhibits good performance for detection of oil spills with wide swath SAR imagery. The Precision and Recall metrics are 89.23% and 89.14%, respectively. The average Precision is 92.56%. The effects of environmental conditions and sensor parameters on oil spill detection are analyzed. The expected detection results are obtained when wind speeds and incidence angles are between 3 m/s and 10 m/s, and 21° and 45°, respectively. Furthermore, the computer runtime for oil spill detection is less than 0.05 s for each full SAR image, using a workstation with NVIDIA GeForce RTX 3090 GPU. This suggests that the present approach has potential for applications that require fast oil spill detection from spaceborne SAR images.

Huang X ，Zhang B ，Perrie W ，Lu Y ，Wang C ... -  《-》

Ocean oil spill detection from SAR images based on multi-channel deep learning semantic segmentation.

One of the major threats to marine ecosystems is pollution, particularly, that associated with the offshore oil and gas industry. Oil spills occur in the world's oceans every day, either as large-scale spews from drilling-rig or tanker accidents, or as smaller discharges from all sorts of sea-going vessels. In order to contribute to the timely detection and monitoring of oil spills over the oceans, we propose a new Multi-channel Deep Neural Network (M-DNN) segmentation model and a new and effective Synthetic Aperture Radar (SAR) image dataset, that enable us to emit forewarnings in a prompt and reliable manner. Our proposed M-DNN is a pixel-level segmentation model intended to improve previous DNN oil-spill detection models, by taking into account multiple input channels, complex oil shapes at different scales (dimensions) and evolution in time, and look-alikes from low wind speed conditions. Our methodology consists of the following components: 1) New Multi-channel SAR Image Database Development; 2) Multi-Channel DNN Model based on U-net and ResNet; and 3) Multi-channel DNN Training and Transfer Learning. Due to the lack of public oil spill databases guaranteeing a correct learning process of the M-DNN, we developed our own database consisting of 16 ENVISAT-ASAR images acquired over the Gulf of Mexico during the Deepwater Horizon (DWH) blowout, off the west coast of South Korea during the Hebei Spirit oil tanker collision, and over the Black Sea. These images were pre-processed to create a 3-channel input image IM = {IO, IW, IV}, to feed in and train our M-DNN. The first channel IO represents the radiometric values of the original SAR Images, the second and third channels are derived from IO; in particular, IW represents the output of the wind speed estimation using CMOD5 algorithm (Hersbach et al., 2003) and IV represents the variance of IO that incorporates texture information and at the same time encapsulates oil spill transition regions. IM channels were split and linearly transformed for data augmentation (rotation and reflection) to obtain a total of 80,772 sub-images of 224 × 224 pixels. From the entire database, 80 % of the sub-images were used in the DNN training process, the remaining (20 %) was used for testing our final architecture. Our experimental results show higher pixel-level classification accuracy when 2 or 3 channels are used in the M-DNN, reaching an accuracy of 98.56 % (the highest score reported in the literature for DNN models). Additionally, our M-DNN model provides fast training convergence rate (about 14 times better on the average than previous works), which proves the effectiveness of our proposed method. According to our knowledge, our work is the first multi-channel DNN based scheme for the classification of oil spills at different scales.

Hasimoto-Beltran R ，Canul-Ku M ，Díaz Méndez GM ，Ocampo-Torres FJ ，Esquivel-Trava B ... -  《-》

Near real time monitoring of platform sourced pollution using TerraSAR-X over the North Sea.

Continuous operational monitoring by means of remote sensing contributes significantly towards less occurrence of oil spills over European waters however, operational activities show regular occurrence of accidental and deliberate oil spills over the North Sea, particularly from offshore platform installations. Since the areas covered by oil spills are usually large and scattered over the North Sea, satellite remote sensing particularly Synthetic Aperture Radar (SAR) represents an effective tool for operational oil spill detection. This paper describes the development of a semi-automated approach for oil spill detection, optimized for near real time offshore platform sourced pollution monitoring context. Eight feature parameters are extracted from each segmented dark spot. The classification algorithm is based on artificial neural network. An initial evaluation of this methodology has been carried out on 156 TerraSAR-X images. Wind and current history information also have been analyzed for particular cases in order to evaluate their influences on spill trajectory.

Singha S ，Velotto D ，Lehner S 《-》