Search

Quantifying the kinematic evolution patterns of mountain glaciers near Yarlung Tsanpo River performs a major role in evaluating the glacial instability and the secondary disasters. For the Sedongpu Basin near the Yarlung Tsanpo River Valley, the dramatic geomorphic landscape variations triggered by the ice-rock avalanche events were visually identified as the dominant texture deficiencies in time-series optical images. To improve the image correlation quality broken by these image texture deficiencies, the Landsat-8/Sentinel-2 optical images were divided into different groups, then a stepwise combination strategy was innovatively proposed to derive the glacier time-series displacement velocities in different temporal stages. The standard deviations (STD) of the optical measurements in the stable area maintained around 0.04 m/yr for the normalized displacement velocity and maintained from 0.6 to 1.7 m for the cumulative displacement time series. The obvious variations in glacier displacement velocity were identified before each collapse event. Subsequently, the offset-tracking procedures were performed on 7 Sentinel-1A Synthetic Aperture Radar (SAR) images to acquire the range and azimuth displacement velocities. To better reveal the dynamic mechanism of the glacier activity, the three-dimensional (3D) glacial displacement velocity was also derived by using optical and SAR results. The precipitation, temperature, and seismic activities were assumed as the main triggering factors of controlling the glacial dynamic mechanism and final collapse events. Additionally, the dynamic mechanism of the active glaciers in Sedongpu Basin conformed to a power law, which was limited by the changes of the internal stress friction force on the sliding base surface. The aim of this study is to shed a light on interpreting the precursory displacement patterns and their implicit failure mechanism of these ice-rock avalanche events with the conventional freely optical and SAR observations.

Radargrammetry is a useful approach to generate Digital Surface Models (DSMs) and an alternative to InSAR techniques that are subject to temporal or atmospheric decorrelation. Stereo image matching in radargrammetry refers to the process of determining homologous points in two images. The performance of image matching influences the final quality of DSM used for spatial-temporal analysis of landscapes and terrain. In SAR image matching, local matching methods are commonly used but usually produce sparse and inaccurate homologous points adding ambiguity to final products; global or semi-global matching methods are seldom applied even though more accurate and dense homologous points can be yielded. To fill this gap, we propose a hierarchical semi-global matching (SGM) pipeline to reconstruct DSMs in forested and mountainous regions using stereo TerraSAR-X images. In addition, three penalty functions were implemented in the pipeline and evaluated for effectiveness. To make accuracy and efficiency comparisons between our SGM dense matching method and the local matching method, the normalized cross-correlation (NCC) local matching method was also applied to generate DSMs using the same test data. The accuracy of radargrammetric DSMs was validated against an airborne photogrammetric reference DSM and compared with the accuracy of NASA’s 30 m SRTM DEM. The results show the SGM pipeline produces DSMs with height accuracy and computing efficiency that exceeds the SRTM DEM and NCC-derived DSMs. The penalty function adopting the Canny edge detector yields a higher vertical precision than the other two evaluated penalty functions. SGM is a powerful and efficient tool to produce high-quality DSMs using stereo Spaceborne SAR images.

On 17 June 2019, an Ms6.0 earthquake occurred in Changning, Sichuan, China (Changning event), which was the largest earthquake on record within 50 km of the area. It attracted great attention as the area has the largest shale gas production in China as well as significant mineral salt production. Using the Interferometric Synthetic Aperture Radar (InSAR), we extract the coseismic deformation of the Changning event and two earlier Ms > 5.0 earthquakes which occurred in the same region (16 December 2018 Ms5.7 and 3 January 2019 Ms5.3) from the Sentinel-1 and ALOS2 data. We use nonlinear and linear methods to invert the fault models of the three earthquakes based on the deformation fields. The final model shows that the Changning event was caused by a fault with left-lateral strike and thrust slip. The strike is 124.3° with a dip angle of 43.4°. The seismic moment obtained by inversion is 5.28 × 1017 Nm, corresponding to Mw 5.78. Based on the fault models, we analyze the cause of the Changning earthquake considering the local tectonic setting, Coulomb stress change, mining, and fluid injection. We consider that the event may be related to salt mining. The two earlier Ms > 5.0 earthquakes may also play an important role in advancing the Changning earthquake.

This paper reports on the activities carried out in the context of “Dragon project 32278: Three- and Four-Dimensional Topographic Measurement and Validation”. The research work was split into three subprojects and encompassed several activities to deliver accurate characterization of targets on land surfaces and deepen the current knowledge on the exploitation of Synthetic Aperture Radar (SAR) data. The goal of Subproject 1 was to validate topographic mapping accuracy of various ESA, TPM, and Chinese satellite system on test sites in the EU and China; define and improve validation methodologies for topographic mapping; and develop and setup test sites for the validation of different surface motion estimation techniques. Subproject 2 focused on the specific case of spatially and temporally decorrelating targets by using multi-baseline interferometric (InSAR) and tomographic (TomoSAR) SAR processing. Research on InSAR led to the development of robust retrieval techniques to estimate target displacement over time. Research on TomoSAR was focused on testing or defining new processing methods for high-resolution 3D imaging of the interior of forests and glaciers and the characterization of their temporal behavior. Subproject 3 was focused on near-real-time motion estimation, considering efficient algorithms for the digestion of new acquisitions and for changes in problem parameterization.

With the development of interferometric synthetic aperture radar (InSAR), the seismic deformation observation density increases sharply. Data down-sampling can effectively reduce the observation density and the computational cost for subsequent researches. Considering the saliency of the deformation field, we introduce a saliency-based quadtree algorithm for down-sampling (SQS). Three simulation experiments show that SQS can effectively distinguish the near-field and far-field deformation, as well as reduce the amount of observation, while keeping the detailed information of the main deformation near the fault. SQS can avoid the interference of far-field local deformation better than the traditional quadtree sampling algorithm (QS), thus obtaining better inversion results. We took the Dingri earthquake on 20 March 2020 as a case study to verify the advantages of SQS in dealing with real earthquake deformation. We obtained the co-seismic deformation from the ascending and descending Sentinel-1 for the Dingri earthquake, using QS and SQS for sampling and inversion separately. The results show the advantages of SQS in data volume reduction, observation distribution, anti-interference of local deformation, and inversion accuracy. Our preferred solution based on SQS shows that the Dingri earthquake was caused by a normal fault slip. The main slip area is 2–5.5 km deep with a maximum slip of 0.68 m. The estimated geodetic moment is 3.14 × 1017 Nm, corresponding to a magnitude of Mw5.63.

Semantic segmentation is an important field for automatic processing of remote sensing image data. Existing algorithms based on Convolution Neural Network (CNN) have made rapid progress, especially the Fully Convolution Network (FCN). However, problems still exist when directly inputting remote sensing images to FCN because the segmentation result of FCN is not fine enough, and it lacks guidance for prior knowledge. To obtain more accurate segmentation results, this paper introduces edge information as prior knowledge into FCN to revise the segmentation results. Specifically, the Edge-FCN network is proposed in this paper, which uses the edge information detected by Holistically Nested Edge Detection (HED) network to correct the FCN segmentation results. The experiment results on ESAR dataset and GID dataset demonstrate the validity of Edge-FCN.

Synthetic Aperture Radar (SAR) Tomography is a technique to provide direct three-dimensional (3D) imaging of the illuminated targets by processing SAR data acquired from different trajectories. In a large part of the literature, 3D imaging is achieved by assuming mono-dimensional (1D) approaches derived from SAR Interferometry, where a vector of pixels from multiple SAR images is transformed into a new vector of pixels representing the vertical profile of scene reflectivity at a given range, azimuth location. However, mono-dimensional approaches are only suited for data acquired from very closely-spaced trajectories, resulting in coarse vertical resolution. In the case of continuous media, such as forests, snow, ice sheets and glaciers, achieving fine vertical resolution is only possible in the presence of largely-spaced trajectories, which involves significant complications concerning the formation of 3D images. The situation gets even more complicated in the presence of irregular trajectories with variable headings, for which the one theoretically exact approach consists of going back to raw SAR data to resolve the targets by 3D back-projection, resulting in a computational burden beyond the capabilities of standard computers. The first aim of this paper is to provide an exhaustive discussion of the conditions under which high-quality tomographic processing can be carried out by assuming a 1D, 2D, or 3D approach to image formation. The case of 3D processing is then further analyzed, and a new processing method is proposed to produce high-quality imaging while largely reducing the computational burden, and without having to process the original raw data. Furthermore, the new method is shown to be easily parallelized and implemented using GPU processing. The analysis is supported by results from numerical simulations as well as from real airborne data from the ESA campaign AlpTomoSAR.

With the rapid development of artificial intelligence, how to take advantage of deep learning and big data to classify polarimetric synthetic aperture radar (PolSAR) imagery is a hot topic in the field of remote sensing. As a key step for PolSAR image classification, feature extraction technology based on target decomposition is relatively mature, and how to extract discriminative spatial features and integrate these features with polarized information to maximize the classification accuracy is the core issue. In this context, this paper proposes a PolSAR image classification algorithm based on fully convolutional networks (FCNs) and a manifold graph embedding model. First, to describe different types of land objects more comprehensively, various polarized features of PolSAR images are extracted through seven kinds of traditional decomposition methods. Afterwards, drawing on transfer learning, the decomposed features are fed into multiple parallel and pre-trained FCN-8s models to learn deep multi-scale spatial features. Feature maps from the last layer of each FCN model are concatenated to obtain spatial polarization features with high dimensions. Then, a manifold graph embedding model is adopted to seek an effective and compact representation for spatially polarized features in a manifold subspace, simultaneously removing redundant information. Finally, a support vector machine (SVM) is selected as the classifier for pixel-level classification in a manifold subspace. Extensive experiments on three PolSAR datasets demonstrate that the proposed algorithm achieves a superior classification performance.

Synthetic Aperture Rradar (SAR) provides rich ground information for remote sensing survey and can be used all time and in all weather conditions. Polarimetric SAR (PolSAR) can further reveal surface scattering difference and improve radar’s application ability. Most existing classification methods for PolSAR imagery are based on manual features, such methods with fixed pattern has poor data adaptability and low feature utilization, if directly input to the classifier. Therefore, combining PolSAR data characteristics and deep network with auto-feature learning ability forms a new breakthrough direction. In fact, feature learning of deep network is to realize function approximation from data to label, through multi-layer accumulation, but finite layers limit the network’s mapping ability. According to manifold hypothesis, high-dimensional data exists in potential low-dimensional manifold and different types of data locates in different manifolds. Manifold learning can model core variables of the target, and separate different data’s manifold as much as possible, so as to complete data classification better. Therefore, taking manifold hypothesis as a starting point, nonlinear manifold learning integrated with fully convolutional networks for PolSAR image classification method is proposed in this paper. Firstly, high-dimensional polarized features are extracted based on scattering matrix and coherence matrix of original PolSAR data, whose compact representation is mined by manifold learning. Meanwhile, drawing on transfer learning, pre-trained Fully Convolutional Networks (FCN) model is utilized to learn deep spatial features of PolSAR imagery. Considering complementary advantages, weighted strategy is adopted to embed manifold representation into deep spatial features, which are input into support vector machine (SVM) classifier for final classification. A series of experiments on three PolSAR datasets have verified effectiveness and superiority of the proposed classification algorithm.

Due to the development of deep convolutional neural networks (CNNs), great progress has been made in semantic segmentation recently. In this paper, we present an end-to-end Bayesian segmentation network based on generative adversarial networks (GANs) for remote sensing images. First, fully convolutional networks (FCNs) and GANs are utilized to realize the derivation of the prior probability and the likelihood to the posterior probability in Bayesian theory. Second, the cross-entropy loss in the FCN serves as an a priori to guide the training of the GAN, so as to avoid the problem of mode collapse during the training process. Third, the generator of the GAN is used as a teachable spatial filter to construct the spatial relationship between each label. Some experiments were performed on two remote sensing datasets, and the results demonstrate that the training of the proposed method is more stable than other GAN based models. The average accuracy and mean intersection (MIoU) of the two datasets were 0.0465 and 0.0821, and 0.0772 and 0.1708 higher than FCN, respectively.

A catastrophic landslide happened on 15 March 2019 in Xiangning County of Shanxi Province, causing 20 fatalities. Such an event makes us realize the significance of loess slope instability detection. Therefore, it is essential to identify the potential active landslides, monitor their displacements, and sort out dominant controlling factors. Synthetic Aperture Radar (SAR) Interferometry (InSAR) has been recognized as an effective tool for geological hazard mapping with wide coverage and high precision. In this study, the time series InSAR analysis method was applied to map the unstable areas in Xiangning County, as well as surrounding areas from C-band Sentinel-1 datasets acquired from March 2017 to 2019. A total number of 597 unstable sites covering 41.7 km2 were identified, among which approximately 70% are located in the mountainous areas which are prone to landslides. In particular, the freezing and thawing cycles might be the primary triggering factor for the failure of the Xiangning landslide. Furthermore, the nonlinear displacements of the active loess slopes within this region were found to be correlated significantly with precipitation. Therefore, a climate-driven displacement model was employed to explore the quantitative relationship between rainfall and nonlinear displacements.

The classification of synthetic aperture radar (SAR) images is of great importance for rapid scene understanding. Recently, convolutional neural networks (CNNs) have been applied to the classification of single-polarized SAR images. However, it is still difficult due to the random and complex spatial patterns lying in SAR images, especially in the case of finite training data. In this paper, a pattern statistics network (PSNet) is proposed to address this problem. PSNet borrows the idea from the statistics and probability theory and explicitly embeds the random nature of SAR images in the representation learning. In the PSNet, both fluctuation and pattern representations are extracted for SAR images. More specifically, the fluctuation representation does not consider the rigorous relationships between local pixels and only describes the average fluctuation of local pixels. By contrast, the pattern representation is devoted to hierarchically capturing the interactions between local pixels, namely, the spatial patterns of SAR images. The proposed PSNet is evaluated on three real SAR data, including spaceborne and airborne data. The experimental results indicate that the fluctuation representation is useful and PSNet achieves superior performance in comparison with related CNN-based and texture-based methods.

Semantic segmentation is an important process of scene recognition with deep learning frameworks achieving state of the art results, thus gaining much attention from the remote sensing community. In this paper, an end-to-end conditional random fields generative adversarial segmentation network is proposed. Three key factors of this algorithm are as follows. First, the network combines generative adversarial network and Bayesian framework to realize the estimation from the prior probability to the posterior probability. Second, the skip connected encoder-decoder network is combined with CRF layer to implement end-to-end network training. Finally, the adversarial loss and the cross-entropy loss guide the training of the segmentation network through back propagation. The experimental results show that our proposed method outperformed FCN in terms of mIoU for 0.0342 and 0.11 on two data sets, respectively.

The convolutional neural network (CNN) has shown great potential in many fields; however, transferring this potential to synthetic aperture radar (SAR) image interpretation is still a challenging task. The coherent imaging mechanism causes the SAR signal to present strong fluctuations, and this randomness property calls for many degrees of freedom (DoFs) for the SAR image description. In this paper, a statistics learning network (SLN) based on the quadratic form is presented. The statistical features are expected to be fitted in the SLN for SAR image representation. (i) Relying on the quadratic form in linear algebra theory, a quadratic primitive is developed to comprehensively learn the elementary statistical features. This primitive is an extension to the convolutional primitive that involves both nonlinear and linear transformations and provides more flexibility in feature extraction. (ii) With the aid of this quadratic primitive, the SLN is proposed for the classification task. In the SLN, different types of statistics of SAR images are automatically extracted for representation. Experimental results on three datasets show that the SLN outperforms a standard CNN and traditional texture-based methods and has potential for SAR image classification.

Thanks to the availability of large-scale data, deep Convolutional Neural Networks (CNNs) have witnessed success in various applications of computer vision. However, the performance of CNNs on Synthetic Aperture Radar (SAR) image classification is unsatisfactory due to the lack of well-labeled SAR data, as well as the differences in imaging mechanisms between SAR images and optical images. Therefore, this paper addresses the problem of SAR image classification by employing the Generative Adversarial Network (GAN) to produce more labeled SAR data. We propose special GANs for generating SAR images to be used in the training process. First, we incorporate the quadratic operation into the GAN, extending the convolution to make the discriminator better represent the SAR data; second, the statistical characteristics of SAR images are integrated into the GAN to make its value function more reasonable; finally, two types of parallel connected GANs are designed, one of which we call PWGAN, combining the Deep Convolutional GAN (DCGAN) and Wasserstein GAN with Gradient Penalty (WGAN-GP) together in the structure, and the other, which we call CNN-PGAN, applying a pre-trained CNN as a discriminator to the parallel GAN. Both PWGAN and CNN-PGAN consist of a number of discriminators and generators according to the number of target categories. Experimental results on the TerraSAR-X single polarization dataset demonstrate the effectiveness of the proposed method.

Persistent Scatterer Interferometry (PS-InSAR) has become an indispensable tool for monitoring surface motion in urban environments. The interferometric configuration of PS-InSAR tends to mix topographic and deformation components in differential interferometric observations. When the upcoming constellation missions such as, e.g., TanDEM-L or TWIN-L provide new standard operating modes, bi-static stacks for deformation monitoring will be more commonly available in the near future. In this paper, we present an analysis of the applicability of such data sets for urban monitoring, using a stack of pursuit monostatic data obtained during the scientific testing phase of the TanDEM-X (TDX) mission. These stacks are characterized by extremely short temporal baselines between the TerraSAR-X (TSX) and TanDEM-X acquisitions at the same interval. We evaluate the advantages of this acquisition mode for urban deformation monitoring with several of the available acquisition pairs. Our proposed method exploits the special properties of this data using a modified processing chain based on the standard PS-InSAR deformation monitoring procedure. We test our approach with a TSX/TDX mono-static pursuit stack over Guangzhou, using both the proposed method and the standard deformation monitoring procedure, and compare the two results. The performance of topographic and deformation estimation is improved by using the proposed processing method, especially regarding high-rise buildings, given the quantitative statistic on temporal coherence, detectable numbers, as well as the PS point density of persistent scatters points, among which the persistent scatter numbers increased by 107.2% and the detectable height span increased by 78% over the standard processing results.

In the areas with steep topography and vulnerable geological condition, landslide deformation monitoring is an important task for risk assessment and management. Differential Synthetic-Aperture Radar interferometry (D-InSAR) and Persistent Scatterer Interferometry (PS-InSAR) are two advanced SAR Interferometry techniques for detection, analysis and monitoring of slow moving landslides. The techniques can be used to identify wide displacement areas and measure displacement rates over long time series with millimeter-level accuracy. In this paper, multiple SAR datasets of Advanced Land Observing Satellite (ALOS) Phased Array L-band Synthetic Aperture Radar (PALSAR) and Environmental Satellite (ENVISAT) C-band Advanced Synthetic Aperture Radar (ASAR) are used for landslide monitoring with both D-InSAR and PS-InSAR techniques in Badong at the Three Gorges area in China. Two areas of significant deformation along the southern riverbank of Yangtze River in Badong are identified by joint analyses of PS-InSAR results from different data stacks. Furthermore, both qualitative and quantitative evaluations of the PS-InSAR results are carried out together with preliminary correlation analysis between the time series deformation of a PS point in high risk location and the temporal variation of water level in the Three Gorges Reservoir.

Automatic image registration of optical-to-Synthetic aperture radar (SAR) images is difficult because of the inconsistency of radiometric and geometric properties between the optical image and the SAR image. The intensity-based methods may require many calculations and be ineffective when there are geometric distortions between these two images. The feature-based methods have high requirements on features, and there are certain challenges in feature extraction and matching. A new automatic optical-to-SAR image registration framework is proposed in this paper. First, modified holistically nested edge detection is employed to detect the main contours in both the optical and SAR images. Second, a mesh grid strategy is presented to perform a coarse-to-fine registration. The coarse registration calculates the feature matching and summarizes the preliminary results for the fine registration process. Finally, moving direct linear transformation is introduced to perform a homography warp to alleviate parallax. The experimental results show the effectiveness and accuracy of our proposed method.

Excessive groundwater exploitation is common through the Taiyuan basin, China, and is well known to result in ground subsidence. However, most ground subsidence studies in this region focus on a single place (Taiyuan city), and thus fail to demonstrate the regional extent of the deformation phenomena in the whole basin. In this study, we used Interferometric Synthetic Aperture Radar (InSAR) time series analysis to investigate land subsidence across the entire Taiyuan basin region. Our data set includes a total of 75 ENVISAT ASAR images from two different frames acquired from August 2003 to September 2010 and 33 TerraSAR-X scenes spanning between March 2009 and March 2010. ERA-Interim reanalysis was used to correct the stratified delay to reduce the bias expected from the systematic components of tropospheric delay. The residual delay after correction of stratified delay can be considered as a stochastic component and be mitigated through spatial-temporal filtering. A comparison with MERIS (Medium-Resolution Imaging Spectrometer) measurements indicates that our atmospheric corrections improved agreement over the conventional spatial-temporal filtering by about 20%. The displacement results from our atmosphere-corrected time series InSAR were further validated with continuous GPS data. We found eight subsiding centers in the basin and a surface uplift to the north of Taiyuan city. The causes of ground deformation are analyzed and discussed in relation to gravity data, pre-existing faults, and types of land use.

Interferometric synthetic aperture radar (InSAR) is an effective technology for generating high-precision digital elevation models (DEMs). However, the vertical accuracy of InSAR DEMs is limited by the contradiction between height measurement sensitivity and phase unwrapping reliability in terms of normal baseline length as well as data voids caused by layover or shadow effects. In order to alleviate these two unfavorable factors, in this study, a novel InSAR DEM fusion method with guided filter is developed and assessed with multiple bistatic TanDEM-X InSAR data pairs of different normal baselines acquired from different orbits. Unlike the widely used fusion method based on pixel-by-pixel weighted average, the guided-filter-based method incorporates local spatial context information into the fusion and can thus effectively alleviate the noise effect and automatically fill in data voids. As a result of the local edge-preserving capability of the guided filter, the proposed fusion method can preserve terrain details by maintaining gradient consistency and introducing terrain features as guidance image. Furthermore, the proposed fusion method is computationally efficient owing to the linear time complexity of guided filter. The experimental results show that the fused DEM with guided filter can depict terrain details well and smooth the “salt-and-pepper” noise and fill in almost all of the data voids. The root mean square error (RMSE) of the fused InSAR DEM with guided filter is lower than those of the weighted average fused InSAR DEM and the TanDEM-X DEM released by the German Aerospace Center (DLR), thus validating the effectiveness of the fusion method proposed in this study.

Temperature variation plays a significant role in the long-term structural behaviour of civil infrastructures, but very few quantitative studies have measured and analysed the infrastructures’ global thermal dilation because of their large sizes and geometric complexities. The modern Differential Synthetic Aperture Radar Interferometry (DInSAR) technique has great potential in applications of their thermal dilation mapping and characterization due to the techniques’ unique capabilities for use in large areas, with high-resolution, and at low-costs for deformation measurements. However, the practical application of DInSAR in thermal dilation estimation is limited by difficulty in the precise separation from the residual topographic phase and the trend deformation phase. Moreover, due to a lack of thermal dilation characteristics analyses in previous studies, the thermal dilation mechanisms are still unclear to users, which restricts the accurate understanding of the thermal dilation evolution process. Given the above challenges, an advanced multi-temporal DInSAR approach is proposed in this study, and the effectiveness of this approach was presented using three cases studies concerning different infrastructure types. In this method, the coherent, incoherent, and semantic information of structures were combined in order to refine the detection of point-like targets (PTs). Interferometric subsets with small temporal baselines and temperature differences were used for high-resolution topography estimation. A pre-analysis was adopted to determine the transmission direction, spatial pattern, and temporal variation of the thermal dilation. Then, both the traditional least squares estimation and our robust coherence-weighted least squares regression analysis were performed between the time series displacements and the corresponding temperatures to quantitatively estimate the thermal dilation model. The results were verified in terms of the estimated linear thermal dilation coefficient, which indicates the greater reliability of our method. Furthermore, the thermal dilation characteristics of different civil infrastructure types were analysed, which facilitates a greater understanding of the thermal dilation evolution process of civil infrastructures.

In this paper, a component-based multi-layer parallel network is proposed for airplane detection in Synthetic Aperture Radar (SAR) imagery. In response to the problems called sparsity and diversity brought by SAR scattering mechanism, depth characteristics and component structure are utilized in the presented algorithm. Compared with traditional features, the depth characteristics have better description ability to deal with diversity. Component information is contributing in detecting complete targets. The proposed algorithm consists of two parallel networks and a constraint layer. First, the component information is introduced into the network by labeling. Then, the overall target and corresponding components are detected by the trained model. In the following discriminative constraint layer, the maximum probability and prior information are adopted to filter out wrong detection. Experiments for several comparative methods are conducted on TerraSAR-X SAR imagery; the results indicate that the proposed network has a higher accuracy for airplane detection.

For InSAR topographic mapping, multi-baseline InSAR height estimation is known to be an effective way to facilitate phase unwrapping by significantly increasing the ambiguity intervals and maintaining good height measurement sensitivity, especially in mountainous areas. In this paper, an efficient multi-baseline SAR interferometry approach based on maximum likelihood estimation is developed for refined topographic mapping in mountainous areas. In the algorithm, maximum likelihood (ML) height estimation is used to measure the topographic details and avoid the complicated phase unwrapping process. In order to be well-adapted to the mountainous terrain conditions, the prior height probability is re-defined to take the local terrain conditions and neighboring height constraint into consideration in the algorithm. In addition, three strategies are used to optimize the maximum likelihood height estimation process to obtain higher computational efficiency, so that this method is more suitable for spaceborne InSAR data. The strategies include substituting a rational function model into the complicated conversion process from candidate height to interferometric phase, discretizing the continuous height likelihood probability, and searching for the maximum likelihood height with a flexible step length. The experiment with simulated data is designed to verify the improvement of the ML height estimation accuracy with the re-defined prior height distribution. Then the optimized processing procedure is tested with the multi-baseline L-band ALOS/PALSAR data covering the Mount Tai area in China. The height accuracy of the generated multi-baseline InSAR DEM can meet both standards of American DTED-2 and Chinese national 1:50,000 DEM (mountain) Level 2.

Inspired by enormous success of fully convolutional network (FCN) in semantic segmentation, as well as the similarity between semantic segmentation and pixel-by-pixel polarimetric synthetic aperture radar (PolSAR) image classification, exploring how to effectively combine the unique polarimetric properties with FCN is a promising attempt at PolSAR image classification. Moreover, recent research shows that sparse and low-rank representations can convey valuable information for classification purposes. Therefore, this paper presents an effective PolSAR image classification scheme, which integrates deep spatial patterns learned automatically by FCN with sparse and low-rank subspace features: (1) a shallow subspace learning based on sparse and low-rank graph embedding is firstly introduced to capture the local and global structures of high-dimensional polarimetric data; (2) a pre-trained deep FCN-8s model is transferred to extract the nonlinear deep multi-scale spatial information of PolSAR image; and (3) the shallow sparse and low-rank subspace features are integrated to boost the discrimination of deep spatial features. Then, the integrated hierarchical subspace features are used for subsequent classification combined with a discriminative model. Extensive experiments on three pieces of real PolSAR data indicate that the proposed method can achieve competitive performance, particularly in the case where the available training samples are limited.

Large-scale reclamation projects during the past decades have been recognized as one of the driving factors behind land subsidence in coastal areas. However, the pattern of temporal evolution in reclamation settlements has rarely been analyzed. In this work, we study the spatio-temporal evolution pattern of Linggang New City (LNC) in Shanghai, China, using space-borne synthetic aperture radar interferometry (InSAR) methods. Three data stacks including 11 X-band TerraSAR-X, 20 L-band ALOS PALSAR, and 35 C-band ENVISAT ASAR images were used to retrieve time series deformation from 2007 to 2010 in the LNC. An InSAR analysis from the three data stacks displays strong agreement in mean deformation rates, with coefficients of determination of about 0.9 and standard deviations for inter-stack differences of less than 4 mm/y. Meanwhile, validations with leveling data indicate that all the three data stacks achieved millimeter-level accuracies. The spatial distribution and temporal evolution of deformation in the LNC as indicated by these InSAR analysis results relates to historical reclamation activities, geological features, and soil mechanisms. This research shows that ground deformation in the LNC after reclamation projects experienced three distinct phases: primary consolidation, a slight rebound, and plateau periods.

This paper proposes an innovative Adaptive Component Selection-Based Discriminative Model (ACSDM) for object detection in high-resolution synthetic aperture radar (SAR) imagery. In order to explore the structural relationships between the target and the components, a multi-scale detector consisting of a root filter and several part filters is established, using Histogram of Oriented Gradient (HOG) features to describe the object from different resolutions. To make the detected components of practical significance, the size and anchor position of each component are determined through statistical methods. When training the root model and the corresponding part models, manual annotation is adopted to label the target in the training set. Besides, a penalty factor is introduced to compensate information loss in preprocessing. In the detection stage, the Small Area-Based Non-Maximum Suppression (SANMS) method is utilised for filtering out duplicate results. In the experiments, the aeroplanes in TerraSAR-X SAR images are detected by the ACSDM algorithm and different comparative methods. The results indicate that the proposed method has a lower false alarm rate and can detect the components accurately.

This paper describes an improved Constant False Alarm Rate (CFAR) ship detection algorithm in spaceborne synthetic aperture radar (SAR) image based on Alphastable distribution model. Typically, the CFAR algorithm uses the Gaussian distribution model to describe statistical characteristics of a SAR image background clutter. However, the Gaussian distribution is only valid for multilook SAR images when several radar looks are averaged. As sea clutter in SAR images shows spiky or heavy-tailed characteristics, the Gaussian distribution often fails to describe background sea clutter. In this study, we replace the Gaussian distribution with the Alpha-stable distribution, which is widely used in impulsive or spiky signal processing, to describe the background sea clutter in SAR images. In our proposed algorithm, an initial step for detecting possible ship targets is employed. Then, similar to the typical two-parameter CFAR algorithm, a local process is applied to the pixel identified as possible target. A RADARSAT-1 image is used to validate this Alpha-stable distribution based algorithm. Meanwhile, known ship location data during the time of RADARSAT-1 SAR image acquisition is used to validate ship detection results. Validation results show improvements of the new CFAR algorithm based on the Alpha-stable distribution over the CFAR algorithm based on the Gaussian distribution.

This paper proposes an innovative Mixture Statistical Distribution Based Multiple Component (MSDMC) model for target detection in high spatial resolution Synthetic Aperture Radar (SAR) images. Traditional detection algorithms usually ignore the spatial relationship among the target’s components. In the presented method, however, both the structural information and the statistical distribution are considered to better recognize the target. Firstly, the method based on compressed sensing reconstruction is used to recover the SAR image. Then, the multiple component model composed of a root filter and some corresponding part filters is applied to describe the structural information of the target. In the following step, mixture statistical distributions are utilised to discriminate the target from the background, and the Method of Logarithmic Cumulants (MoLC) based Expectation Maximization (EM) approach is adopted to estimate the parameters of the mixture statistical distribution model, which will be finally merged into the proposed MSDMC framework together with the multiple component model. In the experiment, the aeroplanes and the electrical power towers in TerraSAR-X SAR images are detected at three spatial resolutions. The results indicate that the presented MSDMC Model has potential for improving the detection performance compared with the state-of-the-art SAR target detection methods.

Contradictions in spatial resolution and temporal coverage emerge from earth observation remote sensing images due to limitations in technology and cost. Therefore, how to combine remote sensing images with low spatial yet high temporal resolution as well as those with high spatial yet low temporal resolution to construct images with both high spatial resolution and high temporal coverage has become an important problem called spatio-temporal fusion problem in both research and practice. A Multi-Dictionary Bayesian Spatio-Temporal Reflectance Fusion Model (MDBFM) has been proposed in this paper. First, multiple dictionaries from regions of different classes are trained. Second, a Bayesian framework is constructed to solve the dictionary selection problem. A pixel-dictionary likehood function and a dictionary-dictionary prior function are constructed under the Bayesian framework. Third, remote sensing images before and after the middle moment are combined to predict images at the middle moment. Diverse shapes and textures information is learned from different landscapes in multi-dictionary learning to help dictionaries capture the distinctions between regions. The Bayesian framework makes full use of the priori information while the input image is classified. The experiments with one simulated dataset and two satellite datasets validate that the MDBFM is highly effective in both subjective and objective evaluation indexes. The results of MDBFM show more precise details and have a higher similarity with real images when dealing with both type changes and phenology changes.

Since the Persistent Scatterer Synthetic Aperture Radar (SAR) Interferometry (PSI) technology allows the detection of ground subsidence with millimeter accuracy, it is becoming one of the most powerful and economical means for health diagnosis of major transportation infrastructures. However, structures of different types may suffer from various levels of localized subsidence due to the different structural characteristics and subsidence mechanisms. Moreover, in the complex urban scenery, some segments of these infrastructures may be sheltered by surrounding buildings in SAR images, obscuring the desirable signals. Therefore, the subsidence characteristics on different types of structures should be discussed separately and the accuracy of persistent scatterers (PSs) should be optimized. In this study, the PSI-based subsidence mapping over the entire transportation network of Shanghai (more than 10,000 km) is illustrated, achieving the city-wide monitoring specifically along the elevated roads, ground highways and underground subways. The precise geolocation and structural characteristics of infrastructures were combined to effectively guide more accurate identification and separation of PSs along the structures. The experimental results from two neighboring TerraSAR-X stacks from 2013 to 2016 were integrated by joint estimating the measurements in the overlapping area, performing large-scale subsidence mapping and were validated by leveling data, showing highly consistent in terms of subsidence velocities and time-series displacements. Spatial-temporal subsidence patterns on each type of infrastructures are strongly dependent on the operational durations and structural characteristics, as well as the variation of the foundation soil layers.

Landslides constitute a major threat to people’s lives and property in mountainous regions such, as in the Three Gorges area in China. Synthetic Aperture Radar Interferometry (InSAR) with its wide coverage and unprecedented displacement measuring capabilities has been widely used in landslide monitoring. However, it is difficult to apply traditional InSAR techniques to investigate landslides having large deformation gradients or moving primarily in north-south direction. In this study, we propose a time series split-bandwidth interferometry (SBI) procedure to measure two dimensional (azimuth and range) displacements of the Shuping landslide in the Three Gorges area with 36 TerraSAR-X high resolution spotlight (HS) images acquired from February 2009 to April 2010. Since the phase based SBI procedure is sensitive to noise, we focused on extracting displacements of corner reflectors (CRs) installed on or surrounding the Shuping landslide. Our results agreed well with measurements obtained by the point-like targets offset tracking (PTOT) technique and in-situ GPS stations. Centimeter level accuracy could be achieved with SBI on CRs which shows great potential in futures studies on fast moving geohazards.

This paper presents a classification approach based on attribute learning for high spatial resolution Synthetic Aperture Radar (SAR) images. To explore the representative and discriminative attributes of SAR images, first, an iterative unsupervised algorithm is designed to cluster in the low-level feature space, where the maximum edge response and the ratio of mean-to-variance are included; a cross-validation step is applied to prevent overfitting. Second, the most discriminative clustering centers are sorted out to construct an attribute dictionary. By resorting to the attribute dictionary, a representation vector describing certain categories in the SAR image can be generated, which in turn is used to perform the classifying task. The experiments conducted on TerraSAR-X images indicate that those learned attributes have strong visual semantics, which are characterized by bright and dark spots, stripes, or their combinations. The classification method based on these learned attributes achieves better results.

Classification using the rich information provided by time-series and polarimetric Synthetic Aperture Radar (SAR) images has attracted much attention. The key point is to effectively reveal the correlation between different dimensions of information and form a joint feature. In this paper, a multi-dimensional SAR descriptive primitive for each single pixel is firstly constructed, which in the polarimetric scale obtains incoherent information through target decompositions while in the time scale obtains coherent information through stochastic walk. Secondly, for the purpose of feature extraction and dimension reduction, a special feature space mapping for the descriptive primitive of the whole image is proposed based on sparse manifold expression and compressed sensing. Finally, the above feature is inputted into a support vector machine (SVM) classifier. This proposed method can inherently integrate the features of polarimetric SAR times series. Experiment results on three real time-series polarimetric SAR data sets show the effectiveness of our presented approach. The idea of a multi-dimensional descriptive primitive as a convenient tool also opens a new spectrum of potential for further processing of polarimetric SAR image time series.

This paper presents a hierarchical classification approach for Synthetic Aperture Radar (SAR) images. The Conditional Random Field (CRF) and Bayesian Network (BN) are employed to incorporate prior knowledge into this approach for facilitating SAR image classification. (1) A multilayer region pyramid is constructed based on multiscale oversegmentation, and then, CRF is used to model the spatial relationships among those extracted regions within each layer of the region pyramid; the boundary prior knowledge is exploited and integrated into the CRF model as a strengthened constraint to improve classification performance near the boundaries. (2) Multilayer BN is applied to establish the causal connections between adjacent layers of the constructed region pyramid, where the classification probabilities of those sub-regions in the lower layer, conditioned on their parents’ regions in the upper layer, are used as adjacent links. More contextual information is taken into account in this framework, which is a benefit to the performance improvement. Several experiments are conducted on real ESAR and TerraSAR data, and the results show that the proposed method achieves better classification accuracy.

Many coastal cities have undertaken reclamation projects to satisfy the land demands of rapid urbanization. However, the foundations of reclaimed land are susceptible to settlement and can have undesirable environmental impacts that could adversely affect these dense, populated areas. In the case of international airports built on reclaimed areas especially, regional-scale deformation must be monitored to ensure operational security for public safety. Persistent Scatterer SAR Interferometry (PSI) technology has proven to be an effective tool to detect ground deformation in urban areas. However, it is still a challenge to apply PSI to effectively monitor settlement at airports built on newly developed coastal reclamation areas because of the scarcity of identifiable targets. Moreover, additional issues arise as the complicated deformation patterns associated with the underlying geological conditions make it difficult to interpret InSAR-derived results. In this study, a time-series analysis of a high-resolution TerraSAR-X satellite image stack acquired from September 2011 to October 2012 was performed by employing a modified PSI technique to retrieve the mean deformation velocity and time series of surface deformation at Pudong International Airport. Qualitative evaluation of spatial distribution and temporal evolution of deformation was conducted by joint analyses of deformation measurements and local geological data. Detailed analysis of various driving forces for deformation patterns confirmed that the results of deformation monitoring obtained by PSI are reliable and consistent with that of local geological surveys. Since the factors responsible for the subsidence within the airport are still at play, ongoing and routine deformation monitoring is warranted.

The Altai Mountains are a heritage-rich archaeological landscape with monuments in almost every valley. Modern nation state borders dissect the region and limit archaeological landscape analysis to intra-national areas of interest. Remote sensing can help to overcome these limitations. Due to its high precision, Synthetic Aperture Radar (SAR) data can be a very useful tool for supporting archaeological prospections, but compared to optical imagery, the detectability of sites of archaeological interest is limited. We analyzed the limitations of SAR using TerraSAR-X images in different modes. Based on ground truth, the discernibility of burial mounds was analyzed in different SAR acquisition modes. We show that very-high-resolution TerraSAR-X staring spotlight images are very well suited for the task, with >75% of the larger mounds being discernible, while in images with a lower spatial resolution only a few large sites can be detected, at rates below 50%.

In recent years, satellite synthetic aperture radar interferometry (InSAR) has been adopted as a spaceborne geodetic tool to successfully measure surface deformation of a few well-known landslides in the Three Gorges area. In consideration of the fact that most events of slope failure happened at places other than those famous landslides since the reservoir impoundment in 2003, focusing on a limited number of slopes is insufficient to meet the requirements of regional-scale landslide disaster prevention and early warning. As a result, it has become a vital task to evaluate the overall stability of slopes across the vast area of Three Gorges using wide-coverage InSAR datasets. In this study, we explored the approach of carrying out joint analysis of multi-path InSAR data stacks for wide-area landslide deformation mapping. As an example, three ALOS (Advanced Land Observing Satellite) PALSAR (Phased Array type L-band Synthetic Aperture Radar) data stacks of neighboring ascending paths covering the area along the Yangtze River from Fengjie to Zigui were analyzed. A key problem to be solved is the separation of the tropospheric signal from the interferometric phase, for which we employed a hybrid description model of the atmospheric phase screen (APS) to improve APS estimation from time series interferograms. The estimated atmospheric phase was largely correlated with the seasonal rainfall in the temporal dimension. The experimental results show that about 30 slopes covering total areas of 48 km2 were identified to be landslides in active deformation and should be kept under routine surveillance. Analyses of time series displacement measurements revealed that most landslides in the mountainous area far away from Yangtze River suffered from linear deformation, whereas landslides located on the river bank were destabilized predominantly by the influences of reservoir water level fluctuation and rainfall.

Time-series SAR/InSAR techniques have proven to be effective tools for measuring landslide movements over large regions. Prior studies of these techniques, however, have focused primarily on technical innovation and applications, leaving coupling analysis of slope displacements and trigging factors as an unexplored area of research. Linking potential landslide inducing factors such as hydrology to SAR/InSAR derived displacements is of crucial importance for understanding landslide deformation mechanisms and could support the development of early-warning systems for disaster mitigation and management. In this study, a sequential data assimilation method named the Ensemble Kalman filter (EnKF), is adopted to explore the response mechanisms of the Shuping landslide movement in relation to hydrological factors. Previous research on the Shuping landslide area shows that the reservoir water level and rainfall are the two main triggering factors in slope failures. To extract the time-series deformations for the Shuping landslide area, Pixel Offset Tracking (POT) technique with corner reflectors was adopted to process the TerraSAR-X StripMap (SM) and High-resolution Spotlight (HS) images. Considering that these triggering factors are the primary causes of displacement fluctuations in periodic displacement, time-series decomposition was carried out to extract the periodic displacement from the POT measurements. The correlations between the periodic displacement and the inducing factors were qualitatively estimated through a grey relational analysis. Based on this analysis, the EnKF method was adopted to explore the response relationships between the displacements and triggering factors. Preliminary results demonstrate the effectiveness of EnKF in studying deformation response mechanisms and understanding landslide development processes.