Complex Synthetic Aperture Radar Data Research Articles

Super-Resolution of Synthetic Aperture Radar Complex Data by Deep-Learning

One of the greatest limitations of Synthetic Aperture Radar imagery is the capability to obtain an arbitrarily high spatial resolution. Indeed, despite optical sensors, this capability is not just limited by the sensor technology. Instead, improving the SAR spatial resolution requires large transmitted bandwidth and relatively long synthetic apertures that for regulatory and practical reasons are impossible to be met. This issue gets particularly relevant when dealing with Stripmap mode acquisitions and with relatively low carrier frequency sensors (where relatively large bandwidth signals are more difficult to be transmitted). To overcome this limitation, in this paper a deep learning based framework is proposed to enhance the spatial resolution of low-resolution SAR images while retaining the complex image accuracy. Results on simulated and real SAR data demonstrate the effectiveness of the proposed framework.

Clutter removal in millimeter wave GB-SAR images using OTSU’s thresholding method

In this study, a filtering method based on the threshold value of normalized synthetic aperture radar (SAR) data is proposed to eliminate clutter in millimeter wave ground based synthetic aperture radar (GB-SAR) images. In the proposed method, first, stepped frequency continuous wave SAR data are reconstructed by using the back-projection algorithm and focused complex SAR data are obtained. Then, the amplitude values of the complex SAR data are normalized and the best threshold values to distinguish the target from clutter is determined by the OTSU’s thresholding method. Next, a filter mask is created that cancels all data below the computed threshold values. The mask matrix is finally multiplied with the resulted GB-SAR data to eliminate all clutter from the image. With the proposed technique, the best threshold value is determined automatically by directly processing the raw data without converting the SAR data into any RGB images. The proposed technique is validated through real GB-SAR experiments that were carried out in the frequency band of 78-81 GHz. In the experiments, challenging GB-SAR data are obtained using high cluttered background materials, and very successful filtering operations are performed with the proposed technique.

Target Recognition in Single-Channel SAR Images Based on the Complex-Valued Convolutional Neural Network With Data Augmentation

In recent years, with the rapid development of deep learning theory, real-valued convolutional neural networks have achieved significant success in the field of synthetic aperture radar (SAR) target recognition. However, different from natural images, the SAR images have complex information due to their special imaging mechanism. Traditional deep learning methods for SAR target recognition only employ the amplitude information and ignore the phase portion, which may sacrifice some useful information in the original complex SAR data. Moreover, the number of samples of SAR images is very limited. This is undoubtedly a huge challenge for the traditional real-valued convolutional neural networks (CNNs) which require numerous labeled data for training. Especially, since the single-channel SAR image contains only one channel, it has less available information than the multiple-channel SAR image. To deal with the above problems, a complex-valued convolutional neural network (CV-Net), for target recognition in single-channel SAR images is proposed in this paper. The amplitude and phase information in the complex SAR data are fully utilized for target recognition. In addition, to alleviate the problem of small samples, this paper also proposed a data augmentation method based on the complex SAR images. The experimental results based on the measured SAR data demonstrate that the proposed algorithm has better performance than the traditional real-valued convolutional neural networks.

Observation and Mitigation of Mutual RFI Between SAR Satellites: A Case Study Between Chinese GaoFen-3 and European Sentinel-1A

In synthetic aperture radar (SAR) data processing, radio frequency interference (RFI) has been recognized as a challenging issue, which can significantly degrade the image quality, including amplitude, phase, and geometry, etc. Most of the RFI sources are direct waves transmitted from the ground. Recently, a new form of RFI, namely the mutual RFI (MRFI), which is a kind of scatter-wave RFI originating from the nearby ground area simultaneously illuminated by different SAR satellites, has been reported. In this paper, the signatures of MRFI are characterized, and an example case study between Chinese GaoFen-3 (GF-3) and European Sentinel-1A (S-1A) is analyzed. In order to remove the artifacts caused by MRFI on SAR images, a novel RFI detector based on spectrum energy cancellation (SEC), which has the capability of detecting MRFI, is developed. Two methods, i.e., the traditional notch filtering method and an improved eigen-subspace projection (ESP) method proposed in this paper, are employed to mitigate MRFI. The former method has robust mitigation performance with a fast processing speed, whereas the latter has an improved mitigation accuracy as well as lower side-lobes for strong scatterers. The methods are designed for cases in which the MRFI has a different radar centre frequency than the useful signal, and some of the data is free from MRFI. The conditions required for the proper use of the approach are discussed. In contrast to the conventional mitigating methods performed on the raw data domain, the proposed approach begins with the focused single look complex (SLC) SAR data. The effectiveness of the proposed approach is demonstrated on several GF-3 SLC SAR images, which were contaminated by MRFI from the S-1A.

Integrating the Reconstructed Scattering Center Feature Maps With Deep CNN Feature Maps for Automatic SAR Target Recognition

Automatic target recognition has been one of the hottest research in synthetic aperture radar (SAR) data processing. Noticing that popular recognition methods cannot utilize multiple features of SAR complex data, a method fused scattering center feature and deep convolutional neural network (CNN) feature is proposed in this letter. This method contains three key parts, namely, scattering center extraction and reconstruction block, CNN feature extraction block, and final feature fusion and classification block. In this process, the scattering center feature and CNN feature are fused at the level of feature maps, which retain the space information of 2-D feature maps. What is more, the proposed half end-to-end strategy realizes the automatic update of weighting parameters in feature extraction network and subnetwork, which promotes a better recognition efficiency. Experimental results on measured SAR data show that the proposed method can achieve better accuracy than other single feature-based methods and feature fusion methods.

An Adaptive and Adjustable Maximum-Likelihood Estimator for SAR Change Detection

Synthetic aperture radar (SAR) change detection can be broadly classified into two categories: noncoherent intensity change detection and coherent change detection (CCD). The former methods, belonging to the field of image processing, are theoretically suitable for all SAR images since only utilize the information of SAR magnitude, however, the detection precision cannot be guaranteed. The latter methods can show effective performances for most SAR image pairs using identical collection geometrics because of the basis of the probability and statistics theory. In this paper, we propose a novel change estimator to combine the advantages of the two kinds of algorithms in a theoretical derivation way. An intensity-based estimator, inspired by the derivation of the coherence estimator in CCD, is first proposed. It is a new maximum-likelihood (ML) change estimator maximizing the probability distribution function (pdf) of the ratio change statistic instead of SAR complex data. In addition, the simple linear iterative cluster (SLIC) algorithm is introduced to make the new estimator adjustable because changes at different degrees can be extracted with varying settings of the superpixels number, which is further demonstrated in real SAR images. Finally, experiments in SAR image pairs of different statistical characteristics show that the proposed estimator can yield higher-contrast SAR change detection images than the other five common change statistics and obtain better change detection maps than the other four classic thresholding methods.

A joint change detection method on complex-valued polarimetric synthetic aperture radar images based on feature fusion and similarity learning

ABSTRACT Existing deep learning-based change detection methods in the field of polarimetric synthetic aperture radar (POLSAR) usually directly deal with intensity images. Methods can be easily transferred from optical image processing to synthetic aperture radar (SAR) image processing. However, the polarization information, which is critical in POLSAR data, has been discarded under normal situations. This paper introduces a novel joint change detection network based on similarity learning for coregistered complex-valued SAR data. A pseudo-siamese network takes both amplitude information and polarization information of POLSAR data as the input. The fusion of low-level, middle-level and high-level features enables the network to keep high-resolution and have strong representation ability during training procedures. Our novel sub-networks, which we term C2-Net and Intensity-Net, deal with the covariance matrix of complex SAR data and amplitude SAR data, respectively. The Intensity-Net works as a typical classification network and detects targets directly. The C2-Net attempts to find the relationship between two SAR data patches. An improved cosine similarity function is used to measure the similarity between two generated feature vectors in C2-Net. Output probability vectors of the two sub-networks are combined for final change detection. The two sub-networks are trained jointly and simultaneously. Control experiments show that proposed improvements are working. Experimental results on dual-polarization complex-valued SAR data from Sentinel-1 demonstrate the effectiveness of the proposed method.

SAR Interference Suppression Algorithm Based on Low-Rank and Sparse Matrix Decomposition in Time–Frequency Domain

Radio frequency electromagnetic interference is a relatively common phenomenon, especially for synthetic aperture radar (SAR) systems working in P- or L-band. Compared with the suppression of narrowband interference, that of wideband interference, particularly of those whose signal parameters have frequently changing property, is still a sophisticated problem. In this letter, a suppression algorithm for interference with wideband and complicated parameters is proposed, based on low-rank and sparse matrix decomposition (LRSMD) in time–frequency domain (TFD) of the signal. The proposed algorithm begins with transforming the SAR signal into TFD. After that, LRSMD based on bilateral random projection (BRP) is applied to decompose the time–frequency spectrum matrix into three parts, a low-rank matrix standing for interference, a sparse matrix standing for SAR signal, and a noise matrix. Finally, inversely transform the sparse matrix into the time domain to obtain SAR signal without interference. The proposed algorithm is applied to a single look complex (SLC) SAR data of Sentinel-1 to validate its effect and efficiency.

FEC: A Feature Fusion Framework for SAR Target Recognition Based on Electromagnetic Scattering Features and Deep CNN Features

The active recognition of interesting targets has been a vital issue for synthetic aperture radar (SAR) systems. The SAR recognition methods are mainly grouped as follows: extracting image features from the target amplitude image or matching the testing samples with the template ones according to the scattering centers extracted from the target complex data. For amplitude image-based methods, convolutional neural networks (CNNs) achieve nearly the highest accuracy for images acquired under standard operating conditions (SOCs), while scattering center feature-based methods achieve steady performance for images acquired under extended operating conditions (EOCs). To achieve target recognition with good performance under both SOCs and EOCs, a feature fusion framework (FEC) based on scattering center features and deep CNN features is proposed for the first time. For the scattering center features, we first extract the attributed scattering centers (ASCs) from the input SAR complex data, then we construct a bag of visual words from these scattering centers, and finally, we transform the extracted parameter sets into feature vectors with the k-means. For the CNN, we propose a modified VGGNet, which can not only extract powerful features from amplitude images but also achieve state-of-the-art recognition accuracy. For the feature fusion, discrimination correlation analysis (DCA) is introduced to the FEC framework, which not only maximizes the correlation between the CNN and ASCs but also decorrelates the features belonging to different categories within each feature set. Experiments on Moving and Stationary Target Acquisition and Recognition (MSTAR) database demonstrate that the proposed FEC achieves superior effectiveness and robustness under both SOCs and EOCs.

Impact of Orthorectification on Simulated Compact Polarimetric RCM Data With Accurate Lidar DSM

Orthorectification using digital terrain models is a key issue for polarimetric complex synthetic aperture radar (SAR) data because resampling the complex data can corrupt the polarimetric phase, mainly in terrain with relief. Orthorectification will be also a sensitive issue with the compact polarimetric data of the future Canadian Radar Constellation Mission (RCM) sensor to be launched in 2018. Two orthorectification methods for the complex SAR data are thus proposed and compared: performing polarimetric processing in the image space before the geometric processing or in the ground space after the geometric processing. RCM compact polarimetric data using the requirements of the very high resolution (VHR) mode were simulated at the Canada Centre for Mapping and Earth Observation from fine-quad Radarsat-2 data acquired with different look angles over a hilly relief study site. Quantitative evaluations between the two methods, using a basis-invariant parameter (the entropy), were thus performed to evaluate the impact of orthorectification on the simulated VHR RCM data. To avoid the propagation of elevation errors into the final error budget, an accurate lidar digital surface model was used in the orthorectification. The results demonstrated that the oversampling and the noise floor that is used to generate the simulated VHR RCM data are the main factors, which corrupted the simulated VHR RCM data during its orthorectification with the ground-space method.

Edge Detection Using Real and Imaginary Decomposition of SAR Data

The objective of synthetic aperture radar (SAR) edge detection is the identification of contours across the investigated scene, exploiting SAR complex data. Edge detectors available in the literature exploit singularly amplitude and interferometric phase information, looking for reflectivity or height difference between neighboring pixels, respectively. Recently, more performing detectors based on the joint processing of amplitude and interferometric phase data have been presented. In this paper, we propose a novel approach based on the exploitation of real and imaginary parts of single-look complex acquired data. The technique is developed in the framework of stochastic estimation theory, exploiting Markov random fields. Compared to available edge detectors, the technique proposed in this paper shows useful advantages in terms of model complexity, phase artifact robustness, and scenario applicability. Experimental results on both simulated and real TerraSAR-X and COSMO-SkyMed data show the interesting performances and the overall effectiveness of the proposed method.

Impact of DEM source on Radarsat-2 polarimetric information during ortho-rectification

Ortho-rectification using digital terrain models is a key issue for full polarimetric complex synthetic aperture radar (SAR) data because resampling the complex data can corrupt the polarimetric phase, mainly in terrain with relief. Two methods for ortho-rectification of the complex SAR data can be applied: the polarimetric processing is performed before (image-space method) or after (ground-space method) the geometric processing. This research evaluated the impact of the digital elevation models (DEMs), which are generally available to users (topographic DEM and ASTER GDEM V2). The two methods were applied to three Radarsat-2 fine-quad data acquired with different look angles over a hilly relief study site. Quantitative evaluations between the two approaches as a function of different geometric and radiometric parameters were, thus, performed to evaluate the impact during the ortho-rectification. The results demonstrated that the look angles and the terrain slopes can potentially corrupt the single-look polarimetric complex SAR data during its ortho-rectification with the ground-space method, mainly at the layover limit. However, advice is provided to reduce these impacts to an acceptable level.

Ocean Vector Winds Retrieval From C-Band Fully Polarimetric SAR Measurements

We present an efficient algorithm for retrieving the ocean-surface wind vector from C-band Radar Satellite RADARSAT-2 fully polarimetric synthetic aperture radar (SAR) measurements based upon the copolarized geophysical model function, i.e., CMOD5.N, and the cross-polarized ocean backscatter model, i.e., C-2PO. The analysis of fine quad-polarization mode single-look complex SAR data and collocated in situ moored buoy observations reveals that the polarimetric correlation coefficient between co- and cross-polarization channels has odd symmetry with respect to the wind direction. This characteristic is different from the feature that normalized radar cross sections for quad-polarization have even symmetry regarding the wind direction. We first use the C-2PO model to directly retrieve wind speeds without any external wind-direction and radar-incidence-angle inputs. Subsequently, the retrieved wind speeds, along with incidence angles and CMOD5.N, are employed to invert the wind direction, still with ambiguities. The odd-symmetry property is then applied to remove the wind direction ambiguities. Thus, it is shown that fully polarimetric SAR measurements provide complementary directional information for the ocean-surface wind fields. This method has the potential to improve wind vector retrievals from space.

Measurement of 2-D Sea Surface Elevation Fields Using Complex Synthetic Aperture Radar Data

Measurement of 2-D Sea Surface Elevation Fields Using Complex Synthetic Aperture Radar Data

Measurement of 2-D sea surface elevation fields using complex synthetic aperture radar data

A method is presented to derive two-dimensional sea surface elevation fields from complex synthetic aperture radar (SAR) data. Applied to spaceborne SAR data as acquired by European Remote Sensing 2 (ERS-2) or the Environmental Satellite (ENVISAT), the method allows to analyze the structure of ocean wave fields, e.g., wave grouping or individual wave heights on a global scale. The technique, thus, provides wave parameters not obtained with common SAR wave retrieval schemes, which are designed to estimate the 2-D wave spectrum, i.e., second-order statistical moments of the wave field. Estimates of sea surface elevation fields are obtained based on the existing theory of SAR ocean wave imaging, i.e., the modulation of the SAR image intensity due real aperture radar and motion-related effects. A power series expansion is derived for SAR intensity images that enables the analysis of nonlinear effects as well as to derive a quasi-linear approximation of the SAR imaging model in the spatial domain. A statistical analysis is performed based on a global dataset of 2D wave spectra provided by the European Centre for Medium-Range Weather Forecast. Distributions are given for the relative error of the quasi-linear approximation in the spatial domain. It is shown that the error can be reduced by smoothing the SAR image in the azimuthal direction at the cost of lower resolution. Smoothed elevation fields are retrieved by the minimization of a cost function defined in the Fourier domain based on the quasi-linear approximation of the imaging process. A multilook technique is applied to infer the information on wave propagation directions, which is required because the SAR transfer function is non-Hermitian, i.e., the SAR image is not determined by the frozen sea surface, but wave motion has a significant impact. The method is applied to simulated SAR images as well as to data acquired by ERS-2. The errors of the retrieved wave field due to image noise, uncertainties in the SAR imaging model, and bandwidth limitations are analyzed. In particular, the fact that the estimated elevation field is smoothed due to the finite system resolution and smearing effects associated with wave motion is discussed. A statistical test is proposed to check the homogeneity of the SAR image. The method makes sure that atmospheric effects are not misinterpreted as being caused by ocean waves.

Velocity and strain rates derived from InSAR analysis over the Amery Ice Shelf, East Antarctica

AbstractWe use displacements derived from matching complex synthetic aperture radar data using maximum coherence to generate a dense network of velocity estimates over the Amery Ice Shelf. From these velocities we generate the horizontal strain-rate components and resolve them with respect to the local flow direction. We present the spatial distributions of velocity and transverse shear strain rate and use them to investigate features of the flow regime for the shelf. From the southern end of the shelf, velocity decreases from a high of about 800ma–1 to around 300 ma–1, and then increases to a maximum of about 1350ma–1 at the centre of the front. Strain rates vary systematically across and along the shelf. The pattern of the transverse shear strain rate clearly identifies the shear margins, where values exceed 0.1 a–1 in the southern section of the shelf. The pattern also shows longitudinal bands of enhanced shear strain rate containing ice with a strong preferred crystal fabric that was advected from shear margins upstream. In the northern section of the shelf, significant values of longitudinal and traverse stresses lead to enhanced shear deformation through their effect on the octahedral shear stress term.

Unsupervised segmentation of polarimetric SAR data using the covariance matrix

A method for unsupervised segmentation of polarimetric synthetic aperture radar (SAR) data into classes of homogeneous microwave polarimetric backscatter characteristics is presented. Classes of polarimetric backscatter are selected on the basis of a multidimensional fuzzy clustering of the logarithm of the parameters composing the polarimetric covariance matrix. The clustering procedure uses both polarimetric amplitude and phase information, is adapted to the presence of image speckle, and does not require an arbitrary weighting of the different polarimetric channels; it also provides a partitioning of each data sample used for clustering into multiple clusters. Given the classes of polarimetric backscatter, the entire image is classified using a maximum a posteriori polarimetric classifier. Four-look polarimetric SAR complex data of lava flows and of sea ice acquired by the NASA/JPL airborne polarimetric radar (AIRSAR) are segmented using this technique.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Segmentation of polarimetric synthetic aperture radar data

A statistical image model is proposed for segmenting polarimetric synthetic aperture radar (SAR) data into regions of homogeneous and similar polarimetric backscatter characteristics. A model for the conditional distribution of the polarimetric complex data is combined with a Markov random field representation for the distribution of the region labels to obtain the posterior distribution. Optimal region labeling of the data is then defined as maximizing the posterior distribution of the region labels given the polarimetric SAR complex data (maximum a posteriori (MAP) estimate). Two procedures for selecting the characteristics of the regions are then discussed. Results using real multilook polarimetric SAR complex data are given to illustrate the potential of the two selection procedures and evaluate the performance of the MAP segmentation technique. It is also shown that dual polarization SAR data can yield segmentation resultS similar to those obtained with fully polarimetric SAR data.

Segmentation of synthetic-aperture-radar complex data

Several algorithms used for segmenting highly speckled high-resolution synthetic-aperture-radar (SAR) complex data into spatially and radiometrically homogeneous regions are presented. The procedure is based on two models, one for the speckled complex amplitudes and the other for the regions. The first model uses the physics of the SAR imaging and processing system to characterize the statistics of speckle, while the second model uses a Markov random field to describe the statistics of the regions. Based on the combination of these two models from Bayes theory, two possible optimality criteria are considered for the segmentation of the complex data into regions. The different algorithms are implemented on a parallel optimization network. Results from both simulated and actual SAR complex data are presented for a comparison of the different alternatives and evaluation of the performance of the segmentation techniques.