Spaceborne synthetic aperture radar (SAR) is an important technology for ship detection applications. It can provide timely information on small boat locations, which is important for security and safety applications. This letter describes the advantages of using very-high-resolution TerraSAR-X data acquired in staring spotlight mode for detecting small boats. Coincident to the SAR image acquisitions, electro-optical (EO) satellite imagery was used, together with field photographs of boats. The results demonstrate the ability to distinguish SAR signatures of small wooden and fiberglass vessels with the size of up to 4 m in length.

The inland water bodies are critical components of ecosystems and hydrologic cycles. Thus, the water extent data are crucially important for hydrological and ecological studies. Due to its high temporal resolution, the Cyclone Global Navigation Satellite System (CYGNSS) has the potential for real-time inland water monitoring. In this letter, a high-resolution machine learning (ML) method for detecting inland water content using the CYGNSS data is implemented via the random undersampling boosted (RUSBoost) algorithm. The CYGNSS data of the year 2018 over the Congo and Amazon basins are gridded into$0.01^{\circ }\, \times \, 0.01^{\circ }$cells. The RUSBoost-based classifier is trained and tested with the CYGNSS data over the Congo basin. The data of the Amazon basin that is unknown to the classifier are then used for further evaluation. By only using the observables extracted from the CYGNSS data, the proposed technique is able to detect 95.4% and 93.3% of the water bodies over the Congo and Amazon basins, respectively. The performance of the RUSBoost-based classifier is also compared with an image processing-based inland water detection method. For the Congo and Amazon basins, the RUSBoost-based classifier has a 3.9% and 14.2% higher water detection accuracy, respectively.

This article investigates and compares polarimetric signatures of icebergs embedded in sea ice and icebergs in open water. The main objective is to study on the backscatter properties of melting iceberg and to check on whether there is any distinguishable property in them in the case of different background clutter conditions (i.e. sea ice and open water). This study results will improve the potential of iceberg detection using radar polarimetry. RADARSAT-2 images have been used for the analysis acquired over locations near the coastline (approximately 3–35 km) of the island of Newfoundland. For analysis, polarimetry parameters, such as co-(HH) and cross-(HV) polarization and several popular decomposition techniques, specifically Pauli, Freeman–Durden, Yamaguchi, Cloud–Pottier, and van Zyl, have been used to determine the polarimetric signatures of icebergs and sea ice. The statistical hypothesis T-test has been applied to achieve a precise comparison among backscatters from different icebergs groups. Statistical results tend to show a dominant surface scattering mechanism for icebergs in all types of clutter conditions. Moreover, icebergs in open water produce larger volume scatter than icebergs in sea ice, whereas icebergs in sea ice produce larger surface scatter than icebergs in open water.

Ship detection and classification in very high resolution (VHR) EO/IR satellite imagery, as primary objectives, were investigated using multiple techniques. Automated algorithms were developed and their performance was evaluated using different satellite image sources (Pleiades, WorldView-2/3). Performance of ship detection algorithms based on traditional (thresholding and saliency) techniques reached probability of detection 80% for low false alarm rates. Deep learning techniques based on convolutional neural networks (CNNs) (YOLOv4 and Mask R-CNN) achieved average precision of 94–95% with 3% of false positives without the need of accurate land and cloud masking. Mask R-CNN also allows accurate determining ship size parameters. The problem of ship and non-ship classification was investigated using traditional and CNN based techniques. Linear Discriminant Analysis, Support Vector Machines and combined classifiers achieved classification accuracies close to 80–90%. At the same time, the usage of a technique based on GoogleNet CNN achieved 99% classification accuracy for ship, small boats and background targets.

Summary Ships and icebergs are similar in size and intensity in SAR images, so it is difficult to distinguish them in remote sensing images. Deep learning is a technique based on neural networks, which has played an important role in image information processing. In order to address the challenge of ship and iceberg classification, we present a convolutional neural network (CNN) based classification method for iceberg and ship discrimination from Sentinel‐1 SAR images with different polarizations and incidence angles. The method is based on the fixed constant false alarm rate (CFAR) detector and the CNN model has three input channels, then the model was trained using parallel algorithm. The CNN is trained using 1443 images and tested using 161 images. The CNN model is also compared with support vector machine (SVM) and k nearest neighbors (kNN) using the same dataset. Comparison shows the CNN‐based method performs the best, and it achieved a validation accuracy of 96%.

This letter presents the validation of an electromagnetic (EM) backscatter model of icebergs at C-band by comparing the performances of target classifiers trained with both modeled and real synthetic aperture radar (SAR) data. Simulated SAR data were obtained in a combination of imaging beam modes and scene parameters to produce 216 simulated Sentinel-1 C-band SAR images. Parameters consisted of Sentinel-1 IW1 (33.1°) and IW3 (43.1°) beam modes with varying wind speed (5 and 10 m/s), wind direction (0°, 45°, and 90°), and target orientation (0°, 45°, and 90°). Simulations were created from an EM SAR simulator called GRECOSAR, which took 3-D profiles of iceberg and ship targets and parameters necessary to closely mimic the real scenes. 3-D models of three icebergs were captured in a field study off the coast of Bonavista, Newfoundland, and Labrador, Canada in June 2017. Three generic ship models were sourced from an online inventory and scaled to a size equivalent to that of the iceberg targets. Real SAR image data were drawn from in-house data set collected from a complementary research program. Classifiers including support vector machine (SVM), Random Forest (RanFor), k-nearest neighbor (kNN), and neural network (NN) were trained with targets from modeled SAR data and then gradually mixed with real SAR data. Target classifier performance from the modeled target data was shown to be similar to classifiers trained entirely from real SAR data. The similarity in accuracy provides an indication of the validity of the modeled SAR data for this specific application.

In this paper, the advantages of using an interferometric method for detecting and characterizing icebergs in sea ice were demonstrated. Iceberg topography was analyzed using single-pass TanDEM-X interferometric synthetic aperture radar (InSAR) data. Multiple InSAR data sets in bistatic mode were acquired over icebergs in sea ice in the Arctic region. InSAR processing was used to extract 3-D elevation information of the sea ice surface. The results firmly demonstrate the capability of TanDEM-X data to characterize the shape of icebergs. Very high resolution (VHR) optical satellite data were collected by Pleiades 1A over the same area to derive digital elevation models (DEMs) of ice features for validation. The accuracy of the extracted topography over icebergs was evaluated by comparing InSAR and optical DEMs. The quantitative comparison demonstrated good correspondence between InSAR and optical DEMs with root-mean-square value values of 2.2 m for icebergs and 0.6 m for sea ice, respectively. Using DEMs derived from VHR optical imagery, it was possible to calculate receiver operating characteristics (ROC) for detecting icebergs in sea using InSAR. The resulting ROC analysis illustrates a good detection performance.

Recent research has confirmed that satellite altimetry can be used for detecting icebergs. In an effort to validate the altimetry-based approach, this study used 105 samples of icebergs contained in both satellite altimeter data and ENVISAT-ASAR scenes for the Weddell Sea area. The probability of detecting icebergs larger than 150 m waterline-length was 47%. The problem of discriminating ships and icebergs based on altimeter measurements was addressed using an ensemble of automated classifiers. A total of ten features were defined from the altimetry signal to be used as predictor variables in supervised classification. The classifier ensemble comprised discriminant functions, k-nearest neighbor, neural networks, support vector machines, and decision tree analysis. Several algorithms successfully classified objects as ships or icebergs with an accuracy exceeding 85%.

Information on icebergs and ice islands is important for climate science and for various marine operations in Arctic and Antarctic. This work investigates capabilities of RADARSAT-2 polarimetric data for detection of icebergs in sea ice. Several iceberg detectors were analyzed with the full polarimetric data acquired in Fine Quad and Fine Quad Wide modes. The results of iceberg detection were validated with the information extracted from very high and medium resolution electro-optical satellite data including stereo datasets. It was demonstrated that the accuracy of iceberg detection depends on polarimetric bands, parameters of detection algorithm and sea ice types. The novelty of the work also includes a demonstration of detection characteristics including false alarm rates. Improvement of detection results for icebergs in pack ice was achieved using Pauli decomposition components, span and advancing detection algorithm.

Mapping iceberg locations and geometrical parameters is important for marine operational applications and climate science. The innovative TanDEM-X mission (TDM) was used for 3D mapping of icebergs in sea ice with the single-pass SAR interferometry (InSAR) method. The extracted digital elevation model (DEM) from TDM InSAR data over icebergs in sea ice was compared to a DEM generated from very-high-resolution (VHR) electro-optical stereo data. The comparison demonstrated a good correspondence between the electro-optical and InSAR-derived DEMs. A significant decrease in root-mean-squared error (RMSE) was achieved after applying spatial filtering. The resulting RMSE for the areas with selected icebergs in sea ice was 2 m.

Mapping linear disturbances from oil and gas extraction and mining in Western Canada traditionally relies on manual digitizing from very high resolution remote sensing data, which usually limits results to small operational area. In this paper, we present an approach of mapping linear disturbances based on time series data analysis of regularly acquired, and low cost satellite data of moderate resolution. This approach involves three steps: line detection based on a multi-scale directional template, line updating based on reappearance frequency and line connection using Hough Transform. This automatic method has been tested over three sites in Alberta, Canada by detecting linear disturbances occurring over the period of 1984-2013 using Landsat imagery. The results show that the method can extract very narrow linear features, including seismic lines, roads and pipelines with a good performance.

This paper reports on the design and implementation of a high resolution low-power synthetic aperture radar (SAR). Our system utilizes the linear frequency-modulated continuous wave (LFMCW) concept, and it operates in the S-band. The generic architecture for our design is presented along with a description of the digital SAR processor. Finally, for demonstration purposes, a focused SAR image for a ground-truthed target is provided.

This paper presents a new insight into the nonlinear dynamics in SAR imagery. For extended targets, the conventional radar resolution theory is violated due to the nonlinear phase modulation induced by the dispersive scatterers. A novel algorithm motivated by the Hilbert view for the nonlinear phenomenon is introduced. Our algorithm may be used to not only detect the dispersive scatterers but also to estimate the nonlinear order of the phase modulation. Our results are demonstrated on a representative real-world target chip.

Information on extreme ice features (ridges, icebergs etc.) is important for various marine operations. Satellite synthetic aperture radar (SAR) imagery is capable of monitoring sea ice, identifying and tracking ice features over broad spatial scales. This work investigates possibilities of extreme sea ice features retrieval from various RADARSAT-2 data. Several different beam modes and analysis techniques, such as polarimetric decompositions, were investigated. It was demonstrated that the spatial frequency of sea ice ridges has a very good correlation with the SAR backscatter coefficient. The problem of discriminating glacier ice from sea ice can be resolved by applying Pauli decomposition to full polarimetric data.

Information on the locations and characteristics of extreme sea ice features (such as, hummocks, ridges, stamukhas and icebergs) is important for various marine applications. Imagery acquired by high resolution optical satellites was previously used for qualitative image interpretation to identify various sea ice features and it is especially valuable when detailed ground validation is not available. Current optical satellites, such as GeoEye-1, are able to acquire images with very high resolution of 0.5m. This work addresses the problem of quantitative retrieval of ice feature parameters from very high resolution optical imagery. The developed algorithms facilitate extraction of ice feature height from shadow and derivation of statistical information on ice deformation parameters. Automated processing of GeoEye-1 image demonstrated capabilities of retrieval of ridge frequency and segmentation of rubble fields.

The mapping and monitoring of sea ice regions represents a key application area for spaceborne synthetic aperture radar (SAR) missions. The European Space Agency (ESA) is under-taking the development of Sentinel-1, an imaging radar mission at C-Band for the continuation of SAR operational applications. Monitoring sea ice zones and the arctic environment is one of the major application areas supported by Sentinel-1. Other spaceborne SAR missions are being developed or considered by a number of space organizations. The study described herein focuses on evaluating the contribution of SAR constellations to operational and scientific monitoring of sea ice conditions through the analysis of multi-parameter SAR datasets (airborne, spaceborne) and through the development and exploitation of a simulation tool able to predict sea ice radar signatures as a function of the ice type and condition and of sensor parameters.

The RADARSAT-2 satellite is an advanced C-band synthetic aperture radar (SAR) with a variety of new modes including options for polarization combinations, resolution, and swath width. This paper examines the potential of multi polarization data for detecting and discriminating ship and iceberg targets Data used in this study consist of well validated airborne Convair-580 SAR and spaceborne ASAR HH/HV and HH/VV. In total, the data set used for evaluating detection and discrimination consists of 901 validated iceberg and ship targets. Optimizing target detection is accomplished using receiver operator curves (ROC) as proposed by [6] and discrimination is conducted using a quadratic discriminant (QD) with feature selection based on sequential forward selection (SFS). In general it was found that detection and discrimination improve with more polarimetric information; however, HH/HV and VV/VH only had nominally less discrimination performance than the quad polarization modes evaluated.

Mechanical damage incurred from unauthorized third-party activities remains a leading cause of oil and gas pipeline failure, indicating the need for effective strategies to monitor encroachment over extensive sections of pipeline right-of-ways (ROWs). The purposes of the work discussed in this paper are to evaluate the use of polarimetric spaceborne synthetic aperture radar (SAR) for detecting vehicle targets and discriminating them from false alarms and to lay the foundation for integrating RADARSAT-2 products into the existing encroachment management system (EMS). RADARSAT-2 simulated products were created from Convair-580 imagery collected over Calgary, Canada. Results show that target detection is better for higher resolution data, but discrimination of targets from false alarms is better for dual and quad polarization data due to the increased feature set available that captures more of the scattering behavior. Overall, the Ultra Fine HH and dualpolarization Fine HH/HV or quad-polarization Fine are the recommended modes for an EMS using RADARSAT-2

As part of an ongoing research program in data fusion applied to the multisensor environment for iceberg detection on the east coast of Canada, this paper describes the efforts to use modeling of iceberg movement to perform iceberg detection data fusion. Data fusion and sensor combination are integral parts of a cost effective ice management regime for safe offshore oil and gas recovery programs. This paper addresses data fusion in the form of associations of detections to create tracks from multiple sensor sweeps of an area. The Coupled Ice Ocean (CIO) Model of Iceberg Kinematics from the Canadian Ice Services (CIS) is used as a maneuvering model to the tracking process. The fusion system was tested using iceberg detections from the 2003 ice season. All data were collected in the Grand Banks area of Newfoundland, Canada.

Spaceborne synthetic aperture radar (SAR) can provide wide area and all-weather surveillance for iceberg and ship targets. However, the discrimination between icebergs and ships in SAR imagery, especially in the single polarization imagery that has been available over the past decade, is not always reliable. This is especially true when vessel and iceberg size are on the order of the pixel spacing. Present requirements for ocean surveillance with SAR data include a high detection and classification accuracy due to the necessity of comparable performance with other reconnaissance methods, such as aerial. ENVISAT advanced SAR (ASAR) data offers a potential solution to the iceberg-ship discrimination problem. ASAR data has comparable swath and resolution to other operational SAR systems and in addition offers an alternating polarization (AP) mode. AP targets offer more information than single polarization with respect to radar scattering mechanisms. The AP ship and iceberg targets in this study were observed to have considerably different polarization responses. In particular, ship targets in the HH and HV channels were comparable. In contrast, iceberg targets had at best, weak HV responses compared to the HH channel. Two methods for target discrimination were investigated: a multipolarized area ratio and HV signal-to-clutter ratio (SCR).

SAR satellites can detect icebergs over very large areas in all weather, regardless of ambient conditions like darkness, rain, and fog. As a result, SAR satellites are currently being used on the Grand Banks to enhance iceberg surveillance. During the iceberg season, the Grand Banks region is also frequented by sea ice, which is known to degrade radar performance. To quantify this performance, a validation of the iceberg detection capabilities of satellite radar in sea ice conditions was conducted. The satellite radars used in this validation process were RADARSAT-1 synthetic aperture radar (SAR) and ENVISAT advanced synthetic aperture radar (ASAR). Several sources have documented average radar signature backscatter values for various sea ice types in microwave radar imagery. As part of the validation effort discussed here, sea ice backscatter values were extracted from RADARSAT-1 SAR and ENVTSAT ASAR data and validated against these documented values. Iceberg data for various iceberg sizes have also been extracted from data for both satellites. Using the sea ice backscatter values and the iceberg data information, probability of detection (POD) curves were generated for small, medium and large sized icebergs in multiyear ice, first-year ice, thin lead ice and typical ice conditions found on the Grand Banks. This work was performed with the support of the European Space Agency (ESA), the Canadian Ice Service (CIS) and a consortium of oil and gas companies off the east coast of Newfoundland

Mechanical damage incurred from unauthorized third party activities remains a leading cause of onshore oil and gas pipeline failure, indicating the need for effective strategies to monitor encroachment over extensive sections of pipeline right-of-way (ROW). In This work, the use of polarimetric SAR imagery (as will be available from RADARSAT-2) for pipeline monitoring of encroachment activities is explored. Experimental data were acquired of a test area near the shores of Lake Simcoe (north of Toronto, Ontario) in September 2001 by the C-SAR on board the Convair-580. The vehicle deployments and ground truthing were conducted by C-CORE with processing from signal data (including calibration) and analysis performed at the Canada Centre for Remote Sensing.

Ground movement poses a hazard to buried infrastructure (e.g., pipelines) as either massive instantaneous movement leading to serious failures or as small incremental movement over a period of time resulting in catastrophe, are common. Conventional ground motion detection and measurement techniques require regular costly surveys and operational cost increase in geographically dispersed areas. The measurement of ground movement (e.g., subsidence) using DInSAR (Differential Interferometric Synthetic Aperture Radar) is a well established technique and many studies have been completed to validate its performance. In these studies, ground movement is generally derived from a single look direction (e.g., subsidence only). Strictly speaking, if one look direction is used to obtain ground movement estimates, then only one dimension of movement can be derived. This can produce significant errors in measuring subsidence, especially in the case of localized movement producing significant lateral movement. This work discusses a technique for extracting lateral ground displacement using interferometry and a data fusion, least squares estimation technique from multiple SAR look directions. It is shown that meaningful three dimensional movement estimates can be derived within expected sensor error, with a dramatic reduction in errors of subsidence measurements over a single look direction measurement. The validation was accomplished using a time series of satellite data collected over a pipeline right of way in 2001, along with a series of traditional surveys for comparison.

Regions of extensive marine activity, particularly regions that span hundreds of nautical miles, require wide-area, timely, and reliable remotely sensed information to ensure safe and efficient offshore operations. In some areas, such as the western North Atlantic, this issue is exacerbated for one-third of the year when the region can be frequented by icebergs. This paper reports on development of a framework for "fusing" the various sensor data to provide the most accurate and timely "picture" of the region of interest. Available data includes satellite SAR (synthetic aperture radar) imagery, long-range HF (high frequency) radar, airborne radar, conventional and enhanced marine radar from ships and platforms, and human observation. Work to date has focused primarily on developing performance curves for the various sensors based on empirical data collected over the past two years. This paper presents an overview of that work and the parameters to be optimized in combining data from the disparate sensors.

Since the initial Canadian Space Agency (CSA) ADRO-1 program in 1997, C-CORE has been investigating the capabilities of the RADARSAT synthetic aperture radar (SAR) satellite for the detection of icebergs. This multiyear program has received support from a variety of sources including the CSA's ADRO-1 and ADRO-2 programs, the Canadian Ice Service, and a consortium of oil and gas companies operating on the east coast of Newfoundland. During this program, various RADARSAT modes, including Wide2, Wide3 and ScanSAR Narrow-B have been validated. Threshold and probability of detection curves were generated for small, medium and large sized icebergs in various wind conditions. For these curves, the radar cross-section values from ocean clutter were modeled using the CMOD4 wind model and verified with point source wind measurements. The performance curves show a reasonable success rate for detecting icebergs whose size is on the order of the resolution cell, despite the significant effect of wind speed on detection.

Third party mechanical damage is the principal cause of on-land pipeline failure and is related to such activities as road construction, cable laying, farming and residential and commercial land development. Earth observation data with wide area coverage offers an alternative to the high cost of aerial patrol on large pipeline networks. In order to demonstrate the feasibility of using satellite imagery to detect third party encroachment upon pipeline rights-of-way, two field experiments were carried out in 2000 and 2001. These experiments were conducted at test sites in Alberta and Ontario, Canada. Satellite imagery evaluated during these trials included IKONOS and EROS-A1 optical imagery, RADARSAT-1 fine beam imagery and simulated RADARSAT-2 imagery. Both IKONOS and EROS-A1 imagery were effective in detecting a range of vehicles in areas free of cloud cover. RADARSAT-1 images were effective in the detection of larger vehicles independent of atmospheric conditions. Using simulated RADARSAT-2 it was possible to detect smaller vehicles not previously detected with RADARSAT-1 imagery.