X-band SAR Images Research Articles

Bayesian Data Fusion for Enhanced Monitoring of Bridge Displacements Using Satellite InSAR and Topographic Techniques

This paper investigates the effectiveness of satellite Interferometric Synthetic Aperture Radar (InSAR) technology to remotely monitor the structural health of road bridges and detect possible unexpected behaviours. The study is based on an Italian multi-span prestressed concrete bridge and a dataset of eight years of X-band SAR images collected by the COSMO-SkyMed satellite constellation. The method implemented for this study is as follows. First, SAR images are analysed using the Multi-Temporal InSAR technique to extract displacement time series of reflective targets on the bridge decks. Then, the extracted displacements are interpreted to validate their consistency with the bridge's expected responses to environmental loads, such as air temperature and river water flow variations, and detect possible unexpected behaviours. Eventually, the correlation with the air temperature is investigated to identify different bridge spans based only on monitoring data. The findings from this study advance the knowledge on using satellite InSAR-based monitoring for early detection of anomalous bridge behaviours on a large scale, providing valuable insights for the maintenance and safety of critical infrastructure.

Intricate multiple scattering features of artificial facilities in X-Band SAR images

Abstract. Due to the intricate distortion and reflection geometry of the SAR signal, it is typically difficult to determine the multiple scattering of large artificial objects in SAR images. This work presents a scattering point path tracking model that utilizes the real three-dimensional dimensions of targets, based on the geometric optics method. Three different artificial structures, including light poles, cable stayed bridges, and power transmission lines, are carefully analysed in time-series SAR images with their simulated multiple scattering results. The results demonstrate that the routes determined by the model are consistent with the multiple scattering features on SAR images. Moreover, the time-series data demonstrate that ripples in the water's surface have a significant impact on the multi-scattering features of power lines and bridges. The double scattering features of the light pole provides a novel approach to the process of permanent scatterers (PS) in urban areas. The instances presented in this study demonstrate the effectiveness of the scattering point path tracking model in identifying the various artificial facility targets on different reflective surfaces. It will be a useful tool for deciphering the multiple scattering of large artificial structures when their 3D model is known.

Combination of Time Series of L-, C-, and X-Band SAR Images for Land Cover and Crop Classification

The availability of new Earth observation satellites operating radar sensors at different frequencies enables the combination of multiple dimensions of the data (time, frequency, polarimetry and interferometry) in many applications. Image classification is expected to benefit from the diversity of observation. This work illustrates classification experiments carried out with series of images acquired by ALOS-2 PALSAR (L-band), Sentinel-1 (C-band) and TanDEM-X (X-band) in two application domains: land cover classification and crop-type mapping. Their usage, both separately and in combination, serves to identify the complementarity of information. In this work we propose a new colour representation of the pair-wise class separability in the case of using three frequency bands, which help identify which bands (or combinations of them) provide the best performance. Results in terms of accuracy scores (overall and class-specific) show that the use of the three frequency bands always outperforms the individual bands and their pairs. In addition, for both land classification and crop-type mapping the accuracy of using coherence time series is lower than the one obtained with the intensity time series, but there is complementarity in terms of sensitivity when both coherence and intensity time series are used together. The classes which are most benefited at each particular case of study have been identified. Finally, a partial trade-off has been found between the use of multiple frequency bands and the length of the available time series.

A Comparative Assessment of Machine-Learning Techniques for Forest Degradation Caused by Selective Logging in an Amazon Region Using Multitemporal X-Band SAR Images

The near-real-time detection of selective logging in tropical forests is essential to support actions for reducing CO2 emissions and for monitoring timber extraction from forest concessions in tropical regions. Current operating systems rely on optical data that are constrained by persistent cloud-cover conditions in tropical regions. Synthetic aperture radar data represent an alternative to this technical constraint. This study aimed to evaluate the performance of three machine learning algorithms applied to multitemporal pairs of COSMO-SkyMed images to detect timber exploitation in a forest concession located in the Jamari National Forest, Rondônia State, Brazilian Amazon. The studied algorithms included random forest (RF), AdaBoost (AB), and multilayer perceptron artificial neural network (MLP-ANN). The geographical coordinates (latitude and longitude) of logged trees and the LiDAR point clouds before and after selective logging were used as ground truths. The best results were obtained when the MLP-ANN was applied with 50 neurons in the hidden layer, using the ReLu activation function and SGD weight optimizer, presenting 88% accuracy both for the pair of images used for training (images acquired in June and October) of the network and in the generalization test, applied on a second dataset (images acquired in January and June). This study showed that X-band SAR images processed by applying machine learning techniques can be accurately used for detecting selective logging activities in the Brazilian Amazon.

Despeckling of SAR Images Using Shrinkage of Two-Dimensional Discrete Orthonormal S-Transform

Speckles are inherent to SAR. They hide and undermine several relevant information contained in the SAR images. In this paper, a despeckling algorithm using the shrinkage of two-dimensional discrete orthonormal S-transform (2D-DOST) coefficients in the transform domain along with shock filter is proposed. Also, an attempt has been made as a post-processing step to preserve the edges and other details while removing the speckle. The proposed strategy involves decomposing the SAR image into low and high-frequency components and processing them separately. A shock filter is used to smooth out the small variations in low-frequency components, and the high-frequency components are treated with a shrinkage of 2D-DOST coefficients. The edges, for enhancement, are detected using a ratio-based edge detection algorithm. The proposed method is tested, verified, and compared with some well-known models on C-band and X-band SAR images. A detailed experimental analysis is illustrated.

Comparison of Input Image Dimensions for Ship Detection from KOMPSAT-5 SAR Image Using Deep Neural Network

Park, W.; Baek, W.-K.; Won, J.-S., and Jung, H.-S., 2020. Comparison of input image size for ship detection from KOMPSAT-5 SAR image using deep neural networks (DNNs). In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 208-217. Coconut Creek (Florida), ISSN 0749-0208.This study shows the results of ship detection in KOMPSAT-5 X-band SAR images using Deep Learning (DL) based the fast Deep Neural Network (DNN) method with the iterative kernel-based false alram detection algorithm. In addition, this study verifies the detection accuracy according to the size of the input data for are typical errors in SAR images such as backscattering noise, side-lobe, and ghost effect. The test results of the Fast DNN method confirmed that the error caused by the back scattering noise and side-lobe decreased as the input data size increased, however, the error for the ghost effect did not improve significantly. Also, the False Alarm Rate (FAR) was greatly improved using the iterative kernel-based false alram detection algorithm. However, caution is needed when detecting small ships because very small boats disappear. The results of this study showed that the size of the input data was very effective at 51 × 51 or higher when the fast DNN method and the iterative kernel-based false alram detection algorithm were applied to the KOMPSAT-5 images to detect ships, allowing the FAR and Intersaction over Union (IoU) showing relatively high accuracy at 0.270 and 0.611.

Cosmo-SkyMed and TerraSAR-X datasets for geomorphological mapping in the eastern of Marajó Island, Amazon coast

The Amazon coast is marked by the high discharge of sediments and freshwater, macrotidal influence, a wide continental shelf, extensive flood plains and lowered plateaus which make it unique as a delta and estuary landscape. Further, the tropical climate imposes heavy rains and incessant cloudiness that render microwave systems more suitable for cartography. This study proposed to recognize and map the Amazon coastal environments through the X-band Synthetic Aperture Radar, provided by Cosmo-SkyMed (CSK) and TerraSAR-X (TSX) systems. The SAR datasets consisted of interferometric and stereo pairs, restricted to single-revisit and obtained with small interval (1–11 days), under steeper (θ<35°) and shallow (θ≥35°) incidence angles, and during the rainy and dry seasons. From the 4 acquisitions of X-band SAR data, attributes such as the backscattering coefficient, coefficient of variation, texture, coherence, and Digital Surface Model (DSM) were derived, adding each variable in 5 scenarios. These combinations resulted in 20 models, which were submitted individually to the machine learning (ML) classification approach by Random Forest (RF). The backscattering and altimetry described the coastal environments which shared ambiguity and high dispersion, with the lowest separability for vegetated environments such as Mangrove, Vegetated Coastal Plateau and Vegetated Fluvial Marine Terrace. The coherence provided by interferometry was weak (<0.44), even during the dry season, in the other hand, the cross-correlation was strong (>0.60), during the rainy and dry season showing more suitability for radargrammetry on the Amazon coast. The RF models resulted in Kappa coefficient between 0.39 to 0.70, indicating that the use of X-band SAR images at an incidence angle greater than 44° and obtained in the dry season is more appropriated for the morphological mapping. The RF models given by TSX achieved the higher mapping accuracies per scenario of SAR products, in order of 0.48 to 0.63. Despite this, the best classification was carried out by 19 RF model with 0.70, provided by CSK in shallow incidence composed by intensity, texture, coherence and stereo DSM. The CSK and TSX data allowed to map the Amazon coast precisely, based on X-band at single polarization, high spatial resolution and revisit, which has demonstrated the support for detailed cartography scale (1:50,000) and frequent updating (monthly up to yearly).

Automatic Ship Detection Using the Artificial Neural Network and Support Vector Machine from X-Band Sar Satellite Images

In this paper, an automatic ship detection method using the artificial neural network (ANN) and support vector machine (SVM) from X-band SAR satellite images is proposed. When using machine learning techniques, the most important points to consider are (i) defining the proper input neurons and (ii) selecting the correct training data. We focused on generating two optimal input data neurons that (i) strengthened ship targets and (ii) mitigated noise effects by image processing techniques, including median filtering, multi-looking, etc. The median filter and multi-look operations were used to reduce the background noise, and the median filter operation was also used to remove ships in an image in order to maximize the difference between the pixel values of ships and the sea. Through the root-mean-square difference calculation, most ship targets, even including small ships, were emphasized in the images. We tested the performance of the proposed method using X-band high-resolution SAR images including COSMO-SkyMed, KOMPSAT-5, and TerraSAR-X images. An intensity difference map and a texture difference map were extracted from the X-band SAR single-look complex (SLC) images, and then, the maps were used as input neurons for the ANN and SVM machine learning techniques. Finally, we created ship-probability maps through the machine learning techniques. To validate the ANN and SVM results, optimal threshold values were obtained by using the statistical approach and then used to identify ships from the ship-probability maps. Consequently, the level of recall achieved was greater than 90% in most cases. This means that the proposed method enables the detection of most ship targets from X-band SAR images with a reduced number of false detections from negative effects.

Study of the Dissimilar Deformation of Huxian, Near the Qinling Piedmont Fault (China), by Multiband InSAR Time-Series Technology

In recent years, small earthquakes around Huxian, near the Qinling Piedmont Fault (China), have attracted much attention from the local government. Due to the blindness to the characteristics of Huxian ground deformation, it is highly desirable to conduct ground deformation surveys in this area, as well as study the relationship between ground deformation and these earthquakes. To understand the surface-deformation characteristics of Huxian and the seismic background better, InSAR time-series processing was applied to C-, L-, and X-band SAR images of Huxian. InSAR results show that the surface motion was uplifting from 2007 to 2010, and the surface motion subsided from 2014 to 2015. The 2-dimensional deformation rate field from 2007 to 2010 was obtained by using ascending L-band images and descending C-band images. Models in elastic half-space were used to invert the depth of the surface-deformation power source, the results of which are less than 2.0 km below the ground surface. The inversion results were compared with the available seismic records. Possible reasons for the surface deformation are discussed. According to our study, the observed movement has weak correlation with earthquakes even though the small earthquake activity in this area increased suddenly during recent years.

Monitoring Highway Stability in Permafrost Regions with X-band Temporary Scatterers Stacking InSAR.

Interferograms with short wavelength (e.g., X-band) are usually prone to temporal decorrelation in permafrost regions, leading to the unavailability of sufficient high-coherence interferograms for performing conventional time series InSAR analysis. This paper proposes the utilization of temporary scatterers for the stacking InSAR method, thus enabling extraction of subsidence in a permafrost region with limited SAR images and limited high-coherence interferograms. Such method is termed as the temporary scatterers stacking InSAR (TSS-InSAR). Taking the Gonghe-Yushu highway (about 30 km), part of G214 National Highway in Qinghai province (in a permafrost region), as a case study, this TSS-InSAR approach was demonstrated in detail and implemented. With 10 TerraSAR-X images acquired during the period from May 2015 to August 2015, the subsidence along this highway was extracted. In this case the lack of a consistent number of SAR acquisitions limits the possibility to perform other conventional time series InSAR analysis. The results show that the middle part of this highway is in the thermokarst and seasonal frozen soil area, and its accumulated subsidence reach up to 10 cm in 110 days. The thawing phenomena is still the main reason for the instability of highway. The results demonstrate that the TSS-InSAR method can effectively extract the subsidence information in a challenging scenario with limited X-band SAR images and limited high-coherence interferograms, where other time series InSAR-based techniques cannot be applied in a simple way.

RESEARCH ON METHODS OF HIGH COHERENT TARGET EXTRACTION IN URBAN AREA BASED ON PSINSAR TECHNOLOGY

Abstract. PSInSAR technology has been widely applied in ground deformation monitoring. Accurate identification of Persistent Scatterers (PS) is key to the success of PSInSAR data processing. In this paper, the theoretic models and specific algorithms of PS point extraction methods are summarized and the characteristics and applicable conditions of each method, such as Coherence Coefficient Threshold method, Amplitude Threshold method, Dispersion of Amplitude method, Dispersion of Intensity method, are analyzed. Based on the merits and demerits of different methods, an improved method for PS point extraction in urban area is proposed, that uses simultaneously backscattering characteristic, amplitude and phase stability to find PS point in all pixels. Shanghai city is chosen as an example area for checking the improvements of the new method. The results show that the PS points extracted by the new method have high quality, high stability and meet the strong scattering characteristics. Based on these high quality PS points, the deformation rate along the line-of-sight (LOS) in the central urban area of Shanghai is obtained by using 35 COSMO-SkyMed X-band SAR images acquired from 2008 to 2010 and it varies from −14.6 mm/year to 4.9 mm/year. There is a large sedimentation funnel in the cross boundary of Hongkou and Yangpu district with a maximum sedimentation rate of more than 14 mm per year. The obtained ground subsidence rates are also compared with the result of spirit leveling and show good consistent. Our new method for PS point extraction is more reasonable, and can improve the accuracy of the obtained deformation results.

Performance Analysis of Ship Wake Detection on Sentinel-1 SAR Images

A novel technique for ship wake detection has been recently proposed and applied on X-band Synthetic Aperture Radar images provided by COSMO/SkyMed and TerraSAR-X. The approach shows that the vast majority of wake features are correctly detected and validated in critical situations. In this paper, the algorithm was applied to 28 wakes imaged by Sentinel-1 mission with different polarizations and incidence angles with the aim of testing the method’s robustness with reference to radar frequency and resolution. The detection process is properly modified. The results show that the features were correctly classified in 78.5% of cases, whereas false confirmations occur mainly on Kelvin cusps. Finally, the results were compared with the algorithm performance on X-band images, showing that no significant difference arises. In fact, the total false confirmations rate was 15.8% on X-band images and 18.5% on C-band images. Moreover, since the main criticality concerns again the false confirmation of Kelvin cusps, the same empirical criterion suggested for the X-band SAR images yielded a negligible 1.5% of false detection rate.

Fusion of L- and X-band SAR images for DInSAR-based land subsidence monitoring

本論文では，異なる長さの波長を用いて取得された合成開口レーダ（SAR）画像を併用して，高精度に地盤沈下を推定する手法を提案する。差分干渉SAR解析に使用できる画像枚数が少なくなるにつれて，波長が短いSAR画像では位相を復元する際に生じるアンラッピングエラーが発生する可能性が高まる。一方，波長が長いSAR画像では地盤面の小さな変動が検出されにくくなる。提案手法では，まず二種類のSAR画像から個別に変動速度を推定する。次に，類似の変動速度を示す領域のクラスタを個別に生成する。その後，二種類のSAR画像におけるクラスタ間で平均地盤変動速度が一致するように補正を加え，最終的に波長の短いSAR画像によって得られた結果を採用する。提案手法をX，LバンドSARであるTerraSAR-X，PALSAR-２の８枚ずつの画像に適用して，関西国際空港の地盤変動速度を推定した結果，各々の単独解析結果よりも高精度に推定できることが判明した。

Wake Component Detection in X-Band SAR Images for Ship Heading and Velocity Estimation

A new algorithm for ship wake detection is developed with the aim of ship heading and velocity estimation. It exploits the Radon transform and utilizes merit indexes in the intensity domain to validate the detected linear features as real components of the ship wake. Finally, ship velocity is estimated by state-of-the-art techniques of azimuth shift and Kelvin arm wavelength. The algorithm is applied to 13 X-band SAR images from the TerraSAR-X and COSMO/SkyMed missions with different polarization and incidence angles. Results show that the vast majority of wake features are correctly detected and validated also in critical situations, i.e., when multiple wake appearances or dark areas not related to wake features are imaged. The ship route estimations are validated with truth-at-sea in seven cases. Finally, it is also verified that the algorithm does not detect wakes in the surroundings of 10 ships without wake appearances.

Variability of the East Greenland Current in Fram Strait From Subdaily COSMO-SkyMed X-SAR Imagery

Sea ice speed vectors were extracted from COSMO-SkyMed X-band SAR images acquired twice per day from March 2013 to December 2014 over the East Greenland current (EGC) in western Fram Strait. Only winter periods (October-May) were investigated. Maximum ice speed of 81.7 cm/s to SW was detected at 80°N 0°W between November22 and 23, 2013. At 79°N between 7° and 5°W, sea ice speed vectors were mostly directed to S-SW but reversed to the north for short periods. Using ERA-Interim atmospheric data, the southward component of the ice speed (max. +71.7, min. -30.9 cm/s) was correlated (R = 0.93) to the cross-strait sea level pressure gradient. Ice speeds due to the EGC were obtained by subtracting the wind-induced ice speed component estimated by linear regression. Results for winter 2013-2014 show that the mean southward component of the EGC (9.4 cm/s, σ = 3.0) accounted more than half of the mean sea ice speed (17.7 cm/s, σ = 7.2). The EGC oscillated between 5 and 15 cm/s and sustained the southward ice export during short periods when the wind reversed to north. Satellite-based measurements of the surface EGC are the complement to hydrographic measurements in Fram Strait, but further investigation is required to understand how the apparent variability of the current is influenced by the thickness of the ice transiting in the strait or by wind induced north-to-south oscillations of the sea surface in the strait.

Retrieval of Both Soil Moisture and Texture Using TerraSAR-X Images

The aim of this paper is to propose a methodology combing multi-temporal X-band SAR images (TerraSAR-X) with continuous ground thetaprobe measurements, for the retrieval of surface soil moisture and texture at a high spatial resolution. Our analysis is based on seven radar images acquired at a 36° incidence angle in the HH polarization, over a semi-arid site in Tunisia (North Africa). The soil moisture estimations are based on an empirical change detection approach using TerraSAR-X data and ground auxiliary thetaprobe network measurements. Two assumptions were tested: (1) roughness variations during the three-month radar acquisition campaigns were not accounted for; (2) a simple correction for temporal variations in roughness was included. The results reveal a small improvement in the estimation of soil moisture when a correction for temporal variations in roughness is introduced. By considering the estimated temporal dynamics of soil moisture, a methodology is proposed for the retrieval of clay and sand content (expressed as percentages) in soil. Two empirical relationships were established between the mean moisture values retrieved from the seven acquired radar images and the two soil texture components over 36 test fields. Validation of the proposed approach was carried out over a second set of 34 fields, showing that highly accurate clay estimations can be achieved. Maps of soil moisture, clay and sand percentages at the studied site are derived.

A Neural Network-Based Classification for Sea Ice Types on X-Band SAR Images

We examine the performance of an automated sea ice classification algorithm based on TerraSAR-X ScanSAR data. In the first step of our process chain, gray-level co-occurrence matrix(GLCM)-based texture features are extracted from the image. In the second step, these data are fed into an artificial neural network to classify each pixel. Performance of our implementation is examined by utilizing a time series of ScanSAR images in the Western Barents Sea, acquired in spring 2013. The network is trained on the initial image of the time series and then applied to subsequent images. We obtain a reasonable classification accuracy of at least 70% depending on the choice of our ice-type regime, when the incidence angle range of the training data matches that of the classified image. Computational cost of our approach is sufficiently moderate to consider this classification procedure a promising step toward operational, near-realtime ice charting.

Coseismic liquefaction phenomenon analysis by COSMO-SkyMed: 2012 Emilia (Italy) earthquake

The liquefaction phenomenon that occurred in the coseismic phase of the May 20, 2012 Emilia (Italy) earthquake (ML 5.9) is investigated. It was induced by the water pressure increase in the buried and confined sand layers. The level-ground liquefaction was the result of a chaotic ground oscillation caused by the earthquake shaking and the observed failures were due to the upward water flow caused by the excess of pore pressures. We exploited the capability of the differential synthetic aperture radar interferometry (DInSAR) technique to detect soil liquefactions and estimate their surface displacements, as well as the high sensitivity to surface changes of complex coherence, SAR backscattering and intensity correlation. To this aim, a set of four COSMO-SkyMed X-band SAR images, covering the period April 1–June 6, 2012, was used. Geological–geotechnical analysis was also performed in order to ascertain if the detected SAR-based surface effects could be due to the compaction induced by liquefaction of deep sandy layers. In this regards, the results obtained from 13 electrical cone penetrometer tests show the presence of a fine to medium sandy layer at depths, ranging between 9 and 13m, which probably liquefied during the earthquake, inducing vertical displacements between 3 and 16cm. The quantitative results from geological–geotechnical analysis and the surface punctual effects measured by DInSAR are in good agreement, even if some differences are present, probably ascribable to the local thickness and depth variability of the sandy layer, or to lack of deformation detection due to DInSAR decorrelation. The adopted approach permitted us to define the extent of the areas that underwent liquefaction and to quantify the local subsidence related to these phenomena. The latter achievement provides useful information that must be considered in engineering practices, in terms of expected vertical deformations.

Polarimetric Optimization of Temporal Sublook Coherence for DInSAR Applications

The application of differential synthetic aperture radar interferometry (DInSAR) techniques has been traditionally limited to the single-polarimetric case. The launch of satellites with polarimetric capabilities has triggered the synergy between polarimetric and interferometric algorithms, leading to a significant improvement in final DInSAR products. During the last years, the different polarimetric optimization techniques available have been successfully applied to the so-called classical phase quality estimators, i.e., the coherence and the amplitude dispersion estimators. In this context, a new estimator to evaluate the pixels' phase quality, referred to as temporal sublook coherence (TSC), has recently demonstrated to provide promising results in DInSAR applications. The nature of this estimator, which is based on exploiting the spectral properties of pointlike scatterers through the coherence evaluation of different sublooks of the image spectrum, allows its adaptation to the existing polarimetric optimization methods. This letter presents the benefits of extending the TSC estimator to work with fully polarimetric data. For this purpose, a fully polarimetric data set consisting of ten X-band ground-based SAR (GB-SAR) images is employed. The final DInSAR results obtained by means of TSC polarimetric optimization are compared with the ones obtained with its classical single-polarimetric approach, achieving up to more than a twofold increase in the pixels' density.

Flood Mapping With TerraSAR-X in Forested Regions in Estonia

In this study, an extensive flood in Estonia during spring 2010 was mapped with TerraSAR-X data acquired over both open and forested areas. This was the first time when a large scale flooding area was mapped in Estonia by means of spaceborne remote sensing. This was also the first time when X-band SAR images were successfully used for flood mapping under the forest canopy in the temperate forest zone. The tree height in the study region was 15-25 meters on average, and main tree species were birch (leaf-off condition), pine and spruce. The results were compared with ALOS PALSAR and Envisat ASAR images of the same flooding event. In the study area, TerraSAR-X provided on average 3.2 dB higher backscatter over mixed forest flooded areas compared to non-flooded areas. In deciduous and coniferous forests the difference in average backscatter between flooded and non-flooded forests was even greater, 6.2 dB and 4.0 dB, respectively. A supervised classification algorithm was developed to produce high resolution maps of the flooded area from the TerraSAR-X images to demonstrate the flood mapping capability at X-band. Our results show, that spaceborne X-band SAR data, which currently has the highest resolution among the SAR instruments in space, can be used to map floods under forest canopy in temperate zone despite its short wavelength and high attenuation.