Search

In this paper, we propose a total fractional-order variation model for multiplicative noise removal and contrast enhancement of real SAR images. Inspired by the high dynamic intensity range of SAR images, the full content of the SAR images is preserved by normalizing the original data in this model. Then, we propose a degradation model based on the nonlinear transformation to adjust the intensity of image pixel values. With MAP estimator, a corresponding fidelity term is introduced into the model, which is beneficial for contrast enhancement and bias correction in the denoising process. For the regularization term, a gray level indicator is used as a weighted matrix to make the model adaptive. We first apply the scalar auxiliary variable algorithm to solve the proposed model and prove the convergence of the algorithm. By virtue of the discrete Fourier transform (DFT), the model is solved by an iterative scheme in the frequency domain. Experimental results show that the proposed model can enhance the contrast of natural and SAR images while removing multiplicative noise. [ABSTRACT FROM AUTHOR]

The development of remote sensing technology has redefined the approaches to the Earth's surface monitoring. The Copernicus Programme promoted by the European Space Agency (ESA) and the European Union (EU), through the launch of the Synthetic Aperture Radar (SAR) Sentinel-1 and the multispectral Sentinel-2 satellites, has provided a valuable contribution to monitoring the Earth's surface. There are several review articles on the land use/land cover (LULC) matter using Sentinel images, but it lacks a methodical and extensive review in the specific field of land consumption monitoring, concerning the application of SAR images, in particular Sentinel-1 images. In this paper, we explored the potential of Sentinel-1 images to estimate land consumption using mathematical modeling, focusing on innovative approaches. Therefore, this research was structured into three principal steps: (1) searching for appropriate studies, (2) collecting information required from each paper, and (3) discussing and comparing the accuracy of the existing methods to evaluate land consumption and their applied conditions using Sentinel-1 Images. Current research has demonstrated that Sentinel-1 data has the potential for land consumption monitoring around the world, as shown by most of the studies reviewed: the most promising approaches are presented and analyzed. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Remote Sensing is the property of Editorial Office of Journal of Remote Sensing & Science Publishing Co. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Pollution caused by oil spills does irreversible harm to marine biosystems. To find maritime oil spills, Synthetic Aperture Radar (SAR) has emerged as a crucial mean. How to accurately distinguish oil spill areas from other types of areas is a committed step in detecting oil spills. Owing to its capacity to extract multiscale features and its distinctive decoder, the Deeplabv3+ framework has been developed into an excellent deep learning model in field of picture segmentation. However, in some SAR pictures, there is a lack of clarity in the segmentation of oil film edges and incorrect segmentation of small areas. In order to solve these problems, an improved network, named ASA-DRNet, has been proposed. Firstly, a new structure which combines an axial self-attention module with ResNet-18 is proposed as the backbone of DeepLabv3+ encoder. Secondly, a atrous spatial pyramid pooling (ASPP) module is optimized to improve the network's capacity of extracting multiscale features and to increase the speed of model calculation and finally merging low-level features of different resolutions to enhance the competence of network to extract edge information. The experiments show that ASA-DRNet obtains the better results compared to other neural network models. [ABSTRACT FROM AUTHOR]

Although they have achieved great success in optical images, deep convolutional neural networks underperform for ship detection in SAR images because of the lack of color and textual features. In this paper, we propose our framework which integrates prior knowledge into neural networks by means of the attention mechanism. Because the background of ships is mostly water surface or coast, we use clustering algorithms to generate the prior knowledge map from brightness and density features. The prior knowledge map is later resized and fused with convolutional feature maps by the attention mechanism. Our experiments demonstrate that our framework is able to improve various one-stage and two-stage object detection algorithms (Faster R-CNN, RetinaNet, SSD, and YOLOv4) on two benchmark datasets (SSDD, LS-SSDD, and HRSID). [ABSTRACT FROM AUTHOR]

According to existing signatures for various kinds of land cover coming from different spectral bands, i.e., optical, thermal infrared and PolSAR, it is possible to infer about the land cover type having a single decision from each of the spectral bands. Fusing these decisions, it is possible to radically improve the reliability of the decision regarding each pixel, taking into consideration the correlation of the individual decisions of the specific pixel as well as additional information transferred from the pixels’ neighborhood. Different remotely sensed data contribute their own information regarding the characteristics of the materials lying in each separate pixel. Hyperspectral and multispectral images give analytic information regarding the reflectance of each pixel in a very detailed manner. Thermal infrared images give valuable information regarding the temperature of the surface covered by each pixel, which is very important for recording thermal locations in urban regions. Finally, SAR data provide structural and electrical characteristics of each pixel. Combining information from some of these sources further improves the capability for reliable categorization of each pixel. The necessary mathematical background regarding pixel-based classification and decision fusion methods is analytically presented.

Ship detection research has made great progress with deep learning method in recent years. Because some ship targets are too small and the complex scenes are difficult to distinguish from the ship, the accuracy of small-target ship detection in synthetic aperture radar (SAR) images still requires improvement. In order to effectively reduce the interference of complex scenes and locate the small-target more precisely, we propose a model (Dense-YOLOv4-CBAM) based on a densely connected deep neural network with an attention mechanism. The main improvements and contributions of our propose model are three aspects. First, the dense connections is increased in the backbone network (CSP-Darknet53) based on the YOLOv4 framework to enhance the transmission of image features. Second, to refine the transmission of top-down semantic features and bottom-up positioning features in the feature fusion module, we notivatively insert a spatial and channel convolutional attention mechanism (CBAM) into the feature fusion module. This mechanism can reduce the interference in complex scenes and enable a more effective use of each layer's features to reduce the loss of the semantic information of small targets. Finally, we introduce two kinds of accuracy evaluation metrics to evaluate the effectiveness and robustness of the proposed method in extensive experiments. The results show that the proposed model achieves an optimal performance between the accuracy and calculation time compared to other state-of-the-art detection methods on public synthetic aperture radar ship detection datasets. [ABSTRACT FROM AUTHOR]

An oil spill is a major problem for marine life. This spilled oil affects marine life and disturbs the ecosystem. To overcome this problem, SAR sensors have been utilized in marine units to find oil spills because they are free from cloudiness and sun radiation. Even though the oil spill can be detected in SAR images, the processing time and approach are complex. Several techniques have been proposed for the detection of oil spills using matching algorithms, morphological processes, and clustering. But those techniques require multiple processing steps and algorithms for detection. It also depends on user values for detection. Hence, in this paper, the detection of the oil spill in SAR images is automated by utilizing an optimized deep-learning approach. The deep learning approach automatically detects the spilled region using a trained network. Here, the hybrid Moth flame and particle swarm algorithms are used for finding the hyperparameters for training the deep learning network. With this optimized deep learning approach, the oil spill can be detected automatically and its performance verified by implementing the proposed approach in the MATLAB R2020b version. The proposed method's performance is evaluated and compared with the existing technique using accuracy, sensitivity, specificity, Jaccard, and dice. The proposed method produces 99.64% accuracy, which is very high, and only 0.2 classification error, which is very less compared to the existing methods. [ABSTRACT FROM AUTHOR]

The aim of this study is modeling the dependence of maximum velocity of the Trooz Glacier (Kyiv Peninsula, West Antarctica) on air temperature. For this purpose, we processed a time series of meteorological observations at the Akademik Vernadsky station and the ice flow velocity of the Trooz Glacier. The ice velocities were determined from the synthetic aperture radar images, acquired by the Sentinel-1 satellite, for the period from May 2015 to November 2022. The SAR images were processed in the SNAP (Sentinel Application Platform) program using the Offset Tracking method. As a result, 219 ice flow velocity maps were obtained. During the studied period, the maximum velocities varied from 2.64 m/day (August 19, 2015) to 4.05 m/day (April 18, 2020). A functional dependence between the temperature data from the Akademik Vernadsky station and the remote-sensing data on the air temperature above the glacier's surface was established. We combined the three parameters (time series of the maximum velocities of the glacial flow, remote temperature measurements above the glacier, and direct temperature measurements at the Akademik Vernadsky station) in a linear model. In order to increase the accuracy of the modeling, an a posteriori optimization was carried out. As a result, the average error in determining the maximum velocity of the glacier reduced from 23 cm/day to 17 cm/day. [ABSTRACT FROM AUTHOR]

With the development of synthetic aperture radar (SAR) technology, SAR images are being widely applied, such as in SAR image fusion and transformation detection, where image matching is a key procedure. In this paper, a practical and rapid subpixel matching method is proposed for spaceborne SAR images. The subpixel SAR-oriented features from an accelerated segment test (SAR-OFAST) detection operator is employed to improve the detection accuracy of keypoints. By determining the overlapping area of master--slave images, the search range of matching is reduced, and the strategy of block matching is adopted to save computing resources. To verify the accuracy and reliability of the proposed matching method, several groups of SAR images obtained with an identical observation orientation and diverse observation angles were tested using the proposed and other methods. Experimental results demonstrate that the proposed method has a higher speed than other methods with similar performance in other indicators, making it more suitable for scenarios requiring high efficiency. [ABSTRACT FROM AUTHOR]

In this paper we consider and systematize the main significant parameters of natural oil seeps and the possibility for determining them according to radar and optical satellite data. It is shown that, at present, the analysis of time series of optical and radar images makes it possible to determine the main/characteristic parameters of natural oil seeps, making it possible to estimate the location of their underwater sources; the composition, intensity, and volumes of emissions; the seismical situation; and, to some extent, the prospects for the oil and gas potential of the region. The possibilities, limitations, and prospects for using remote sensing data for the considered problems are summarized and discussed. The survey nature of the paper made it possible to consider all modern methods and results of studies of natural oil seeps using remote sensing. [ABSTRACT FROM AUTHOR]

The results of the five-year (2017–2021) monitoring of oil pollution in the Kerch Strait and its prestrait regions are analyzed. The synthetic aperture radar (SAR) images of the European Sentinel-1A and Sentinel-1B satellites, both real-time and archived, representing valuable material for searching, detecting, and studying the character of oil pollution caused by spills of crude oil, petroleum products, and other oily substances, have been used for monitoring. During monitoring, 2597 oil slicks with a total area of 1120.4 km2 are detected on 481 SAR images. The main sources of pollution have been identified, in descending order, they are cargo transshipment areas in the strait, the port of Taman with its berthing complex, and prestrait anchorages in the Sea of Azov and the Black Sea. In addition, natural seep manifestations in the adjacent Black Sea shelf have been discovered, they create a natural hydrocarbon background that should be taken into account. Based on the monitoring results, a tendency towards an overall decrease in spills is revealed, despite their growth in 2016–2019. Monitoring with SAR using the geoinformation approach has shown its high efficiency for the problems of real-time and retrospective analysis of the satellite imagery data concerning oil pollution. [ABSTRACT FROM AUTHOR]

In this study, principal component analysis and k-means clustering (PCAKM) methods for synthetic aperture radar (SAR) data are analyzed to reduce the sensitivity caused by changes in the parameters and input images of the algorithm, increase the accuracy, and make an improvement in the computation time, which are advantageous for scoring in the follow-up plan. Although there are many supervised methods described in the literature, unsupervised methods may be more appropriate in terms of computing time, data scarcity, and explainability in order to supply a trustworthy system. We consider the PCAKM algorithm, which is used as a benchmark method in many studies when making comparisons. Error metrics, computing times, and utility functions are calculated for 22 modified PCAKM regarding difference images and filtering methods. Various images with different characteristics affect the results of the configurations. However, it is evident that the PCAKM becomes less sensitive and more accurate for both the overall results and image results. Scoring by utilizing these results and other map information is a gap and innovation. Obtaining a change map in a fast, explainable, more robust and less sensitive way is one of the aims of our studies on scoring points in the follow-up plan. [ABSTRACT FROM AUTHOR]

In aquaculture, using high-resolution synthetic aperture radar (SAR) images to precisely segment offshore farms is helpful for reasonable layout planning and statistics of breeding density. However, conventional segmentation methods tend to have low accuracy and slow inference speed. Therefore, we propose a novel, precise, and fast segmentation scheme for offshore farms in high-resolution SAR images based on model fusion and half-precision parallel inference. Specifically, we propose several new high-performance improved UNet++ methods and reasonably fuse the test results. At the same time, a simulated annealing strategy and a morphological closing operation are introduced to improve the segmentation accuracy. In addition, we find that resizing the images to 256 × 256 pixels is better than 512 × 512 pixels for this task, which not only has higher segmentation accuracy but can also increase the inference speed by nearly 13%. Furthermore, a novel half-precision parallel inference strategy is proposed, which can fully utilize the GPU and increase the inference speed by 72.6%. Compared with some state-of-the-art methods, the proposed scheme that merges two improved UNet++ achieves superior accuracy with a frequency weighted intersection over union of 0.9876 and a single image inference time of 0.0218 s on the high-resolution SAR offshore farm dataset.

ABSTRACT Change detection based on remote sensing images, has attracted increasing attention from researchers throughout the world. The synthetic aperture radar (SAR) images have become key resources for detecting changes on the land surface. However, due to the presence of speckle noise and its stochastic nature, SAR data require methodologies that consider these peculiarities. This article presents a similarity measure that considers the randomness present in SAR data. To retrieve the random component in the SAR data, we used the stochastic distance. The similarity measure is carefully elaborated as a function of the stochastic distance such that its variation space is the interval [0, 1], facilitating its interpretation. Our proposal shows promising results in two applications: contrast evaluation, ocean surface change detection and binary change map. It is noteworthy that the possible limitations of our proposal are investigated through simulations guided by a Monte Carlo experiment.

Coastal regions and surface waters are among the fundamental biological and social development resources worldwide. For this reason, it is essential to thoroughly monitor these regions to determine and characterize their geographical features and environmental health. These geographical regions, however, present several monitoring challenges when using remotely sensed imagery. Small water bodies tend to be surrounded by swamps, marshes, or vegetation, making accurate border detection difficult. Coastal waters, in turn, experience several phenomena due to winds, undercurrents, and waves, which also hamper the detection of environmental hazards like oil spills. In this work, we propose an automated segmentation algorithm that can be applied to these targets in airborne and spaceborne SAR images. The method is based on pointwise detection in fuzzy borders using a parameter estimation of the G 0 distribution, which has been successfully used in similar contexts. The underlying assumption is that the sought-for border separates regions with different textures, each having different distribution parameters. Then, stochastic distances can identify the most likely point where this parameter change occurs. A curve interpolation algorithm then estimates the actual contour of the body given the detected points. We assess the adequacy of eight stochastic distances that are mostly applied in the literature. We evaluate the performance of our method in terms of similarity between true and detected boundaries on simulated and actual SAR images, achieving promising results. The performance of our proposal is assessed by Hausdorff distance and Intersection over Union. In the case of synthetic data, the selection of the best stochastic distance depends on the parameters of the G I 0 distribution. In contrast, the harmonic-mean and triangular distances produced the best results in detecting borders in three actual SAR images of lagoons. Finally, we present the results of our proposal applied to an image with oil spills using Bhattacharyya, Hellinger, and Jensen–Shannon distances. [ABSTRACT FROM AUTHOR]

Synthetic Aperture Radar (SAR) is the primary equipment used to detect oil slicks on the ocean's surface. On SAR images, oil spill regions, as well as other places impacted by atmospheric and oceanic phenomena such as rain cells, upwellings, and internal waves, appear as dark spots. Dark spot detection is typically the initial stage in the identification of oil spills. Because the identified dark spots are oil slick candidates, the quality of dark spot segmentation will eventually impact the accuracy of oil slick identification. Although certain sophisticated deep learning approaches employing pixels as primary processing units work well in remote sensing image semantic segmentation, finding some dark patches with weak boundaries and small regions from noisy SAR images remains a significant difficulty. In light of the foregoing, this paper proposes a dark spot detection method based on superpixels and deeper graph convolutional networks (SGDCNs), with superpixels serving as processing units. The contours of dark spots can be better detected after superpixel segmentation, and the noise in the SAR image can also be smoothed. Furthermore, features derived from superpixel regions are more robust than those derived from fixed pixel neighborhoods. Using the support vector machine recursive feature elimination (SVM-RFE) feature selection algorithm, we obtain an excellent subset of superpixel features for segmentation to reduce the learning task difficulty. After that, the SAR images are transformed into graphs with superpixels as nodes, which are fed into the deeper graph convolutional neural network for node classification. SGDCN leverages a differentiable aggregation function to aggregate the node and neighbor features to form more advanced features. To validate our method, we manually annotated six typical large-scale SAR images covering the Baltic Sea and constructed a dark spot detection dataset. The experimental results demonstrate that our proposed SGDCN is robust and effective compared with several competitive baselines. This dataset has been made publicly available along with this paper. [ABSTRACT FROM AUTHOR]

Urban road network information is an important part of modern spatial information infrastructure and is crucial for high-precision navigation map production and unmanned driving. Synthetic aperture radar (SAR) is a widely used remote-sensing data source, but the complex structure of road networks and the noises in images make it very difficult to extract road information through SAR images. We developed a new method of extracting road network information from SAR images by considering angular (A) and texture (T) features in the sliding windows and points of interest (POIs, or P), and we named this method ATP-ROAD. ATP-ROAD is a sliding window-based semi-automatic approach that uses the grayscale mean, grayscale variance, and binary segmentation information of SAR images as texture features in each sliding window. Since POIs have much-duplicated information, this study also eliminates duplicated POIs considering distance and then selects a combination of POI linkages by discerning the direction of these POIs to initially determine the road direction. The ATP-ROAD method was applied to three experimental areas in Shanghai to extract the road network using China's Gaofen-3 imagery. The experimental results show that the extracted road network information is relatively complete and matches the actual road conditions, and the result accuracy is high in the three different regions, i.e., 89.57% for Area-I, 96.88% for Area-II, and 92.65% for Area-III. Our method together with our extraction software can be applied to extract information about road networks from SAR images, providing an alternative for enriching the variety of road information. [ABSTRACT FROM AUTHOR]

In agriculture, plants plays a major role and taking attention of plants is very critical. Generally, the plant are affected through various diseases like fungi, virus and bacteria. Finding of these diseases are main challenging task for a plant disease identification and classification. In the past few years, machine learning (ML) methods have been developed for the plant disease detection. But, the advancement in a subsection of ML, that is, DL (deep learning) models provide a great solution in the agricultural areas in the recent decades. The main objective of the paper is to provide the survey of numerous DL classification models for the plant disease detection by analysing the digital, hyper spectral and SAR images. This paper provide the review of different deep learning architectures which is utilized for plant disease identification and classification. The role of digital, hyper spectral and SAR images with deep learning models for plant disease detection is reviewed. Further, the different well-known DL architecture for plant disease classification is studied. In addition, the current challenges and their solutions of plant disease identification are discussed. Also, the application of DL and advantages/disadvantages of DL structure in plant domain are presented. Finally, the future scope of DL structure for plant domain is discussed. The preparation of this review is to permit future research to learn higher competences of deep learning while identifying plant diseases by enhancing system performance and accuracy. [ABSTRACT FROM AUTHOR]

Change detection from synthetic aperture radar images becomes a key technique to detect change area related to some phenomenon as flood and deformation of the earth surface. This paper proposes a transfer learning and Residual Network with 18 layers (ResNet-18) architecture-based method for change detection from two synthetic aperture radar images. Before the application of the proposed technique, batch denoising using convolutional neural network is applied to the two input synthetic aperture radar image for speckle noise reduction. To validate the performance of the proposed method, three known synthetic aperture radar datasets (Ottawa; Mexican and for Taiwan Shimen datasets) are exploited in this paper. The use of these datasets is important because the ground truth is known, and this can be considered as the use of numerical simulation. The detected change image obtained by the proposed method is compared using two image metrics. The first metric is image quality index that measures the similarity ratio between the obtained image and the image of the ground truth, the second metrics is edge preservation index, it measures the performance of the method to preserve edges. Finally, the method is applied to determine the changed area using two Sentinel 1 B synthetic aperture radar images of Eddahbi dam situated in Morocco.

Groundwater dynamics caused by extraction and recharge are one of the primary causes of subsidence in the urban environment. Lahore is the second largest metropolitan city in Pakistan. The rapid expansion of this urban area due to high population density has increased the demand for groundwater to meet commercial and household needs. Land subsidence due to inadequate groundwater extraction has long been a concern in Lahore. This paper aims to present the persistent scatterer interferometry synthetic aperture radar (PS-InSAR) technique for monitoring the recent land subsidence in Lahore, based on the Sentinel-1 data obtained from January 2020 to December 2021. PS-InSAR techniques are very efficient and cost-effective, determining land subsidence and providing useful results. Areas of high groundwater discharge are prone to high subsidence of −110 mm, while the surroundings show an uplifting of +21 mm during the study period. The PS-InSAR study exposes the subsidence area in detail, particularly when the subsoil is characterized by alluvial and clay deposits and large building structures. This type of observation is quite satisfactory and similar to ground-based surface deformation pertinent to a high subsidence rate. Results will enable more effective urban planning, land infrastructure building, and risk assessment related to subsidence. [ABSTRACT FROM AUTHOR]

In point-based heterologous image matching algorithms, high-quality interest point detection directly affects the final image matching quality. In this paper, starting from the detection mechanism of each interest point detector, optical images and SAR images with different resolutions and covering different areas are selected as experimental data. The five state-of-the-art SAR-Harris, UND-Harris, Har-DoG, Harris-Laplace and DoG interest point detectors are analyzed in terms of scale difference adaptability, nonlinear intensity difference adaptability, distribution uniformity, image registration alignment performance and detection efficiency. Then, we performed registration experiments on images from different sensors, at different times, and at different resolutions to further validate our evaluation results. Finally, the applicable image types of each detector are summarized. The experimental results show that SAR-Harris has the best performance in scale difference adaptability, and UND-Harris has the weakest performance. In terms of nonlinear intensity difference adaptability, SAR-Harris and UND-Harris are comparable, and DoG performance is the weakest. The distribution uniformity of UND-Harris is significantly better than other detectors. Although Har-DoG is weaker than Har-Lap and DoG in repeatability, it is better than both in final image alignment performance. DoG is superior in detection efficiency, followed by SAR-Harris. A comprehensive evaluation and a large amount of experimental data are used to evaluate and summarize each detector in detail. This paper provides a useful guide for the selection of interest point detectors during heterologous image matching. [ABSTRACT FROM AUTHOR]

Land subsidence is a major concern in vastly growing metropolitans worldwide. The most serious risks in this scenario are linked to groundwater extraction and urban development. Pakistan's fourth-largest city, Rawalpindi, and its twin Islamabad, located at the northern edge of the Potwar Plateau, are witnessing extensive urban expansion. Groundwater (tube-wells) is residents' primary daily water supply in these metropolitan areas. Unnecessarily pumping and the local inhabitant's excessive demand for groundwater disturb the sub-surface's viability. The Persistent Scatterer Interferometry Synthetic Aperture Radar (PS-InSAR) approach, along with Sentinel-1 Synthetic Aperture Radar (SAR) imagery, were used to track land subsidence in Rawalpindi-Islamabad. The SARPROZ application was used to study a set of Sentinel-1 imagery obtained from January 2019 to June 2021 along descending and ascending orbits to estimate ground subsidence in the Rawalpindi-Islamabad area. The results show a significant increase (−25 to −30 mm/yr) in subsidence from −69 mm/yr in 2019 to −98 mm/yr in 2020. The suggested approach effectively maps, detects, and monitors subsidence-prone terrains and will enable better planning, surface infrastructure building designs, and risk management related to subsidence. [ABSTRACT FROM AUTHOR]

Ocean surface monitoring, emphasizing oil slick detection, has become essential due to its importance for oil exploration and ecosystem risk prevention. Automation is now mandatory since the manual annotation process of oil by photo-interpreters is time-consuming and cannot process the data collected continuously by the available spaceborne sensors. Studies on automatic detection methods mainly focus on Synthetic Aperture Radar (SAR) data exclusively to detect anthropogenic (spills) or natural (seeps) oil slicks, all using limited datasets. The main goal is to maximize the detection of oil slicks of both natures while being robust to other phenomena that generate false alarms, called "lookalikes". To this end, this paper presents the automation of offshore oil slick detection on an extensive database of real and recent oil slick monitoring scenarios, including both types of slicks. It relies on slick annotations performed by expert photo-interpreters on Sentinel-1 SAR data over four years and three areas worldwide. In addition, contextual data such as wind estimates and infrastructure positions are included in the database as they are relevant data for oil detection. The contributions of this paper are: (i) A comparative study of deep learning approaches using SAR data. A semantic and instance segmentation analysis via FC-DenseNet and Mask R-CNN, respectively. (ii) A proposal for Fuse-FC-DenseNet, an extension of FC-DenseNet that fuses heterogeneous SAR and wind speed data for enhanced oil slick segmentation. (iii) An improved set of evaluation metrics dedicated to the task that considers contextual information. (iv) A visual explanation of deep learning predictions based on the SHapley Additive exPlanation (SHAP) method adapted to semantic segmentation. The proposed approach yields a detection performance of up to 94 % of good detection with a false alarm reduction ranging from 14 % to 34 % compared to mono-modal models. These results provide new solutions to improve the detection of natural and anthropogenic oil slicks by providing tools that allow photo-interpreters to work more efficiently on a wide range of marine surfaces to be monitored worldwide. Such a tool will accelerate the oil slick detection task to keep up with the continuous sensor acquisition. This upstream work will allow us to study its possible integration into an industrial production pipeline. In addition, a prediction explanation is proposed, which can be integrated as a step to identify the appropriate methodology for presenting the predictions to the experts and understanding the obtained predictions and their sensitivity to contextual information. Thus it helps them to optimize their way of working. [ABSTRACT FROM AUTHOR]

Oil spill mapping and detection represent a relevant issue from an environmental point of view, given the effects on marine ecosystems. This study presents a new feature space assessment protocol for oil spill mapping using the Google Earth Engine (GEE). First, we selected five free Sentinel-1A sensor images from the GEE catalogue. Next, we processed the features evaluated from Gray Level Co-occurrence Matrix (GLCM) spectral and texture data. A recursive protocol that comprises a sequential classification of the evaluated image was also applied, wherein each iteration, the feature with less importance, was removed based on the Gini index. We used the Random Forest algorithm for image classification. Each image was trained on 10,000 points and evaluated for accuracy, with an equal number of points collected independently. Our results showed that the Sum Average (Savg), Convolution Smooth (Smooth), Cluster Shade Shade, and Gray level Correlation (Corr) features were essential to identify oil spills and increase the accuracy values. The best classification results based on the features removal experiment and global accuracy were Angola (0.9960), Trinidad and Tobago (0.9829), Italy (0.9506), Kuwait (0.9547), and Dubai (0.9344). Furthermore, it revealed that the protocol created was essential for better understanding the parameter space to detect oil spills with SAR images. [ABSTRACT FROM AUTHOR]

In this article, we introduce the 2020 Gaofen Challenge and relevant scientific outcomes. The 2020 Gaofen Challenge is an international competition, which is organized by the China High-Resolution Earth Observation Conference Committee and the Aerospace Information Research Institute, Chinese Academy of Sciences and technically cosponsored by the IEEE Geoscience and Remote Sensing Society and the International Society for Photogrammetry and Remote Sensing. It aims at promoting the academic development of automated high-resolution earth observation image interpretation. Six independent tracks have been organized in this challenge, which cover the challenging problems in the field of object detection and semantic segmentation. With the development of convolutional neural networks, deep-learning-based methods have achieved good performance on image interpretation. In this article, we report the details and the best-performing methods presented so far in the scope of this challenge.

An oriented, connected, acyclic graph can be used to represent a hydraulic network. We can consider the network as a set of arcs and nodes. Our research work entails extracting a drainage network from satellite images by preprocessing the data to have features that will compute the minimum cost function. In this paper, we propose an approach to obtain a drainage network based on information theory and mathematical morphology. Our idea is to consider a drainage network as a set of segments crossed two by two in a node. The best network segments minimize the variance calculated in the directions of propagation, maximize the covered geodesic distance, and are most rectilinear. To get results, we have applied these treatments: filtering, segmentation, minimization of a cost function and mathematical morphology operators. Our proposed approach gives excellent results of high quality and quantity. However, there are still some problems in this study, especially the choice of an adaptive index of linearity that differentiates the drainage network from the surrounding environment. Existing methods use a DEM to extract drainage networks, because they indicate networks where water circulates (3D topography). By contrast, in this study, we use images from ERS-1/2 satellites in order to exploit radar waves, thanks to their penetrating capacity and sensitivity to the dielectric and geometric properties of the land. The temporal and spatial resolutions of the SAR images also allow a historical monitoring of hydraulic networks because of the continual availability of SAR data.

Ship detection for SAR imagery plays a crucial rule in the field of remote-sensing image processing. Superpixel methods have attracted enormous interest in the recent years, resulting in superpixel-based CFAR (Constant False Alarm Rate) ship detection methods for SAR images become a hot research issue. However, existing methods take superpixel generation as a preprocessing, and CFAR detection is essentially carried out at the pixel level, and the influence of speckle noise has not been fundamentally overcome. This paper innovatively proposed a real superpixel-level CFAR detection method, which takes superpixels instead of pixels as the minimum processing primitives on the basis of reliable superpixel segmentation. Firstly, in order to obtain the background superpixel region of each under test superpixel more accurately and efficiently, we designed a ring topology, which was specially developed for non-Euclidean structure data like superpixels. Secondly, the Johnsonsu function is used to fit the superpixel clutters from the background ring, and the detection threshold was obtained by a given false alarm rate. Finally, the candidate targets were further optimized to carry out accurate detection results. Experimental results based on multiple sets of real SAR images shown that our method has significant advantages in terms of detection rate and false alarm rate compared to the traditional K-CFAR and the existing superpixel-based CFAR methods. [ABSTRACT FROM AUTHOR]

Aircraft detection in synthetic aperture radar (SAR) imagery is a challenging task in SAR automatic target recognition (SAR ATR) areas due to aircraft’s extremely discrete appearance, obvious intraclass variation, small size, and serious background’s interference. In this article, a single shot detector (SSD), namely, attentional feature refinement and alignment network (AFRAN), is proposed for detecting aircraft in SAR images with competitive accuracy and speed. Specifically, three significant components, including attention feature fusion module (AFFM), deformable lateral connection module (DLCM), and anchor-guided detection module (ADM), are carefully designed in our method for refining and aligning informative characteristics of aircraft. To represent the characteristics of aircraft with less interference, low-level textural and high-level semantic features of aircraft are fused and refined in AFFM thoroughly. The alignment between aircraft’s discrete backscatting points and convolutional sampling spots is promoted in DLCM. Eventually, the locations of aircraft are predicted precisely in ADM based on aligned features revised by refined anchors. To evaluate the performance of our method, a self-built SAR aircraft sliced dataset and a large scene SAR image are collected. Extensive quantitative and qualitative experiments with detailed analysis illustrate the effectiveness of the three proposed components. Furthermore, the topmost detection accuracy and competitive speed are achieved by our method compared with other domain-specific methods, e.g., dense attention pyramid network (DAPN) and pyramid attention dilated network (PADN), and general convolutional neural network (CNN)-based methods, e.g., Feature Pyramid Network (FPN), Cascade R-CNN, SSD, RefineDet, and RepPoints Detector (RPDet). [ABSTRACT FROM AUTHOR]

The buildings are very complex for detection on SAR images, where the basic features of those are shadows. There are many different representations for SAR shadow. As result it is no possible to use convolutional neural network for building detection directly. In this article we give property analysis of SAR shadows of different type buildings. After that, each region (ROI) prepared for training of building detection is corrected with its own SAR shadow properties. Reconstructions of ROI will be put in a modified YOLO network for building detection with better quality result.

ABSTRACTIn flood hazard assessment, little attention has been given to the seasonal observation of freshwater inundation using Sentinel-1 Synthetic-Aperture Radar (SAR) images in the tropics. To assess these spatial variations, this study examined the inundation magnitudes and reflectance of riparian flora in the raining seasons (February-April and September-November) and dry-periods (May-August and December-January). The inundation areas were determined using an object-oriented classification algorithm, whilst the merits concerning the lake, riparian vegetation and shoreline were well-defined using backscatter-coefficient values. Findings indicated that the SAR images are practical to monitor inundation coverages and discern lake basin specific features such as highland areas, shoreline, water and riparian-vegetation. Seasonally, inundated sizes were comparatively higher in February (902 sq.km) and October (700 sq.km). Backscatter values of inundated parts varied from −16 dB to −19 dB, whereas those of riparian vegetation were higher in the months of May and August. This study hence displays a clear-cut correlation between riparian-vegetation, hydrology and climate variables.

The development of remote sensing technology has redefined the approaches to the Earth’s surface monitoring. The Copernicus Programme promoted by the European Space Agency (ESA) and the European Union (EU), through the launch of the Synthetic Aperture Radar (SAR) Sentinel-1 and the multispectral Sentinel-2 satellites, has provided a valuable contribution to monitoring the Earth’s surface. There are several review articles on the land use/land cover (LULC) matter using Sentinel images, but it lacks a methodical and extensive review in the specific field of land consumption monitoring, concerning the application of SAR images, in particular Sentinel-1 images. In this paper, we explored the potential of Sentinel-1 images to estimate land consumption using mathematical modeling, focusing on innovative approaches. Therefore, this research was structured into three principal steps: (1) searching for appropriate studies, (2) collecting information required from each paper, and (3) discussing and comparing the accuracy of the existing methods to evaluate land consumption and their applied conditions using Sentinel-1 Images. Current research has demonstrated that Sentinel-1 data has the potential for land consumption monitoring around the world, as shown by most of the studies reviewed: the most promising approaches are presented and analyzed.

Copyright of Journal of Signal Processing is the property of Journal of Signal Processing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Speckle filtering is an important step for target detection in SAR images since this effect makes it difficult or even impossible to extract information from these images. There are several filters available in the literature although evaluating their performances is not a trivial task since it requires comparing the filtered images with a speckle-free image, which is generally unknown. This evaluation is even more complex when the features in the images are heterogeneous, for example, from tropical forests. The objective of this study is to evaluate the performance of the Lee, deGrandi, GammaMAP, single Anisotropic Nonlinear Diffusion (ANLD), multitemporal ANLD, Fast Adaptive Nonlocal SAR (FANS), and Fast GPU-Based Enhanced Wiener filters to reduce the speckle present in the COSMO-SkyMed Stripmap X-band images from the Brazilian Amazon forest region. The evaluation was conducted qualitatively through the visual inspection of the ratio image and the edge detection in the ratio images and quantitatively through the $\alpha \beta $ estimator and other statistical parameters of the filtered images. The GammaMAP filter showed the best performances, both qualitatively and quantitatively, and the FANS filter only qualitative. [ABSTRACT FROM AUTHOR]

Although they have achieved great success in optical images, deep convolutional neural networks underperform for ship detection in SAR images because of the lack of color and textual features. In this paper, we propose our framework which integrates prior knowledge into neural networks by means of the attention mechanism. Because the background of ships is mostly water surface or coast, we use clustering algorithms to generate the prior knowledge map from brightness and density features. The prior knowledge map is later resized and fused with convolutional feature maps by the attention mechanism. Our experiments demonstrate that our framework is able to improve various one-stage and two-stage object detection algorithms (Faster R-CNN, RetinaNet, SSD, and YOLOv4) on two benchmark datasets (SSDD, LS-SSDD, and HRSID).

To eliminate the false alarms in the ship target detection effectively for synthetic aperture radar (SAR) images in complex scenes, this article present a novel ship target discrimination algorithm based on bag of words (BOW) model with multiple features and spatial pyramid matching (SPM), which is named MF-SPM-BOW. The proposed discrimination method mainly contains three stages. First, the SAR scale-invariant feature transform (SAR-SIFT) descriptors and gray-level co-occurrence matrix (GLCM) descriptors are extracted as local features to describe the gradient information and texture information of local regions of an image chip. Then, the SPM technique considering its spatial location information-keeping capability is employed to generate global features with excellent discrimination ability. Finally, the support vector machine (SVM) discriminator based on multiple kernel learning is applied to realize feature fusion in image layer and thus identify targets and clutter. Experimental results show that compared with the traditional discrimination methods and the BOW model discrimination methods, the proposed SAR ship target discrimination algorithm achieves better discrimination performance, which can eliminate most of the false alarms in candidate ship target chips effectively.

The synthetic aperture radar (SAR) images are playing an essential role in remote sensing. Various types of internal, external, and environmental noise are affecting the SAR images. Coherent speckle noise is the primary source of noise in SAR images. Such noise can be removed by using a single filter or combination of filters and transform signals. SAR image denoising has been attracting the attention of researchers for the past three decades. The target area and application type are influencing the choice of denoising method. In this paper, the basics of SAR imaging, steps in the pipeline of SAR despeckling process, filters like Lee filter, Frost filter, Kuan Filter and Gamma Maximum a posteriori (MAP) filter, various state of the art despeckling methods and deep learning approaches for SAR despeckling are discussed. Five transforms for despeckling are discussed with literature. The data sets from different radars, the applications, area of importance, and the quality metrics used to evaluate the despeckling quality are discussed in detail that has been available in the literature of the past two decades. [ABSTRACT FROM AUTHOR]

In the last few years, neural networks were introduced to change detection for a better understanding of remote sensing images. However, the designs of these neural networks were time-consuming processes of trial and error, which failed to account for their validity. Thus, a simple and efficient change detection method based on network architecture search in terms of the evolutionary algorithm is proposed to deal with SAR images change detection problems. In the proposed method, an efficient gene encoding is applied to represent the unpredictable optimal depth and the number of neurons in each hidden layer. Besides, a combinatorial evaluation strategy and a self-adaptive network solution selection are designed for effective and reasonable network architectures. What is more, a hidden layer random alignment crossover operator and a drawing lots mutation operator are designed for the enhancement of diversity of network architectures. Experimental results on a few SAR image data sets demonstrate that the proposed method can generate appropriate networks to solving SAR images change detection. [ABSTRACT FROM AUTHOR]

：Objectives: In order to improve the accuracy of SAR （synthetic aperture radar） image change detection, this paper proposes a method of SAR image change detection based on variational method and Markov random field fuzzy local information C‐means clustering （MRFFLICM） method.Methods: Firstly, we fuse the logarithmic ratio images and logarithmic mean ratio images to construct the difference image. Secondly, variational denoising model is established to remove the noise from difference images. Finally, the spatial neighborhood information is introduced into fuzzy local information C ‐means clustering method by using Markov random field to improve the clustering performance.Results: Experiments on two real SAR datasets show that the proposed variational denoising method can avoid removing the small change region and effectively suppress speckle noise of SAR image. Conclusions: The MRFFLICM method can effectively improve the precision of change detection, thus enhancing the adaptability of change detection method. [ABSTRACT FROM AUTHOR]

Speckle noise is an inherent property of SAR images, which will reduce the performance of targets recognition. The contour and shape of targets in speckle noise images may be changed. Although the convolution neural network has been widely used in the field of SAR targets recognition, it learns by automatically extracting images features, the CNN model is still sensitive to noise. In this paper, multiplicative noise is used to simulate speckle noise in SAR images, and the influence of single-source noise and multi-source noise on the recognition performance of CNN is studied. By constructing different noise level data sets, we explore the impact of noise on network performance for three types of MSTAR. Under the premise of ensuring certain target recognition accuracy, we consider that noise can be accepted into the data set for learning, so as to improve the anti-noise performance of intelligent recognition network. [ABSTRACT FROM AUTHOR]

In the Arctic Ocean, the observation of mesoscale eddies is impeded by the presence of sea ice. To address this problem, we develop a new method of ocean eddy detection based on their signature in sea ice vorticity retrieved from synthetic‐aperture radar (SAR) images. We examine the case of one eddy in October 2017 in the marginal ice zone of the Canadian Basin, which was sampled by mooring observations. Although the eddy could not be identified by visual inspection of the SAR images, its signature is revealed as a dipole anomaly in sea ice vorticity, which suggests that the eddy is a dipole composed of a cyclone and an anticyclone, with a horizontal scale of 80–100 km and persisted over a week. The robustness of our method will allow us to detect more eddies as more SAR observations become available in the future. Plain Language Summary: Mesoscale eddies are routinely observed by satellites in the ocean. Yet, in the ice‐covered Arctic Basin, the presence of sea ice makes it challenging to characterize the eddy field. Here, we present a detection method of surface ocean eddies based on their signature in the displacement of sea ice, using high spatial resolution satellite images. A dipole composed of a cyclonic and an anticyclonic eddy is identified over a week in mid‐October 2017 with a horizontal scale of 80–100 km. Its presence is confirmed by high values of ocean speed in the surface layer during the same period. This work demonstrates that processing is required for identifying the signature of eddies in sea ice, which is not always obvious at first sight. Key Points: We develop a new method to identify under‐ice ocean surface eddy in the Arctic from ice vorticity based on synthetic‐aperture radar imagesWe detect the signature of an eddy dipole with a horizontal scale of 80–100 km, persistent on a week, which was also sampled by a mooring [ABSTRACT FROM AUTHOR]

Automatic recognition of ocean eddies has become one of the hotspots in the field of physical oceanography. Traditional methods based on either physical parameters or geometry features require manual parameter adjustment, and cannot adapt to the dynamic changes of ocean eddies caused by complicated ocean environments. To address these issues, we propose a new eddy recognition method in SAR images with adaptive weighted multi-feature fusion. Specially, to better characterize eddies, we first extract texture, shape and corner features using global Gray Level Co-occurrence Matrix (GLCM), detailed Fourier Descriptor (FD) and local salient Harris features respectively. Secondly, considering the different importance of features for eddy recognition, we propose an adaptive weighted feature fusion method with multiple kernel learning (MKL). Here, a combined kernel is derived to fuse three selected kernels for the three types of features with the weights trained by MKL. Finally, we design a SVM classifier with the combined kernel to realize the eddy recognition. The experimental results show that: 1) our proposed method can reach 93.42% of eddy recognition accuracy, which is much higher than the methods with only one single feature; 2) adaptive weighted fusion plays an important role in improving the accuracy. Our proposed method with MKL gains a 8.36% accuracy increase than the method without MKL. Through adaptive weighted fusion, our method avoids the manual parameter adjustment and is more robust and general. Experimental results have proven that our method is effective and applicable to recognize eddies.