Synthetic Aperture Radar Applications Research Articles

The Impact of Variability in SAR Satellite Imagery on Classification

Artificial intelligence (AI) can be a useful tool to gather intelligence from remote sensing data; it helps make sense of synthetic aperture radar (SAR) data via discovery and exploitation. The challenge of utilizing AI in SAR applications is obtaining (large enough) comprehensive sets of labeled training data because SAR data has significant variation across sensor-related characteristics, across processing parameters, and across the different collection plans. This work evaluates the impact of SAR satellite imagery variations on classification accuracy, and demonstrates this by classifying pixels of SAR imagery into land, water, and ship for varying conditions (area-of-interest, incidence angle, spatial resolution, etc.). Results showed that variations in the area-of-interest (AOI), incidence angle, and spatial resolution impacted the classification results obtained using an artificial neural network (ANN). This work also demonstrated that ANNs trained on SAR imagery can be used to infer training data labels of other SAR imagery obtained from different conditions, provided that the changes in condition produced less than a 5% classification error or increased class separation for some (or all) of the classes being discriminated.

Implementation and configuration of GB-SAR for landslide monitoring: case study in Minami-Aso, Kumamoto

ABSTRACTIn this paper, the applicability of ground-based synthetic aperture radar (GB-SAR) as an early warning system for landslide monitoring is discussed. The effectiveness of the differential interferometric SAR (DInSAR) technique used in GB-SAR depends strongly on the geography of the monitored location. Therefore, an assessment of the system compatibility to select the most appropriate remote monitoring method is essential prior to any hardware implementation. In the preliminary part of this study, a 3D model was created using a LiDAR survey, and proposed locations for GB-SAR installation were examined. A 3D simulation was carried out to estimate the illumination from each of the proposed GB-SAR locations. The proposed model increased the efficiency of the GB-SAR positioning by minimising installation cost and time. Hardware configuration parameters, such as platform height, maximum range, and the direction and view angle of the radar line of sight were estimated by considering the optimum reflected power and ground illumination. Unlike on flat terrain, deployment of GB-SAR in a mountainous area is challenging because of surface anomalies and continuous changes in meteorological parameters, such as atmospheric temperature, pressure and relative humidity. In this study, the experimental site was located 3 km from the Aso volcano, and the weather conditions in the Aso caldera became a critical factor in accurately estimating the interferometric phase. The presence of atmospheric artefacts also compromises the applicability of the classical DInSAR technique. Here, we minimised the atmospheric phase screen by estimating the optimum data acquisition interval from GB-SAR monitoring under extreme weather conditions. The developed methodologies were then used to design a new landslide early warning system that measures real-time displacement over an area of 1 km2 within 10 s of scanning. This fully automatic monitoring system updates every 15 min and presents displacement information in a 3D interface. The system we have developed has been deployed for continuous monitoring of the mountainous environment of a road reconstruction site in Minami-Aso, Kumamoto, Japan where a large-scale landslide was triggered following the Kumamoto earthquake in 2016.

Metrics to evaluate compression algorithms for raw SAR data

Modern synthetic aperture radar (SAR) systems have size, weight, power and cost (SWAP-C) limitations since platforms are becoming smaller while SAR operating modes are becoming more complex. Thus, SAR systems are producing an ever-increasing volume of data that needs to be transferred to a ground station for processing. A compression algorithm seeks to reduce the data volume of the raw data; however, the algorithm can cause degradation and losses that may degrade the effectiveness of the SAR mission. This work addresses the lack of standardised quantitative performance metrics so that the performance of SAR data-compression algorithms can be objectively quantified. Therefore, metrics are established in two different domains, namely the data domain and the image domain. Since different levels of degradation are acceptable for different SAR applications, a trade-off can be made between the data reduction and the degradation caused by the algorithm. Due to SWAP-C limitations, there remains a trade-off between the performance and the computational complexity of the compression algorithm.

Parallel Connected Generative Adversarial Network with Quadratic Operation for SAR Image Generation and Application for 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.

The maiden flight of Hinotori-C: The first C band full polarimetric circularly polarized synthetic aperture radar in the world

Synthetic Aperture Radar (SAR) is a very powerful tool in microwave remote sensing due to its capability of all-weather and day-to-night time operation [1]. Carl Wiley invented SAR in 1951 to overcome the poor azimuth resolution in conventional Side-Looking Airborne Radar (SLAR), followed by a patent filed in 1954 [2]. Unlike SLAR, using the forward motion of the platform and the principle of Doppler beam sharpening, the azimuth resolution of a SAR is equal to half the antenna length and is independent of the range distance [3]. Since the invention of SAR, steered by the breakthrough in science and technology, many advanced SAR techniques have been proposed and realized. These techniques include, but are not limited to, spotlight SAR for a finer image resolution [4], scan SAR for a wider swath coverage [5], and the remarkable polarimetric [6] and interferometric [7] SAR techniques using multichannel SAR system [8], [9], [10] for advanced remote sensing applications. Over the years, SAR has been widely used in different types of application, particularly in Earth observation such as disaster damage assessment [11], land deformation observation [12], oceanography [13], terrain classification [14], target detection [15], and so on. The diversified applications of SAR have encouraged the rapid development of airborne and spaceborne SAR sensors.

Ultra‐wide band bowtie slot waveguide antenna array for SAR application

An ultra-wide bandwidth, high-efficiency, small-size and low-profile waveguide antenna is presented in this study. To extend the bandwidth without increasing the antenna size, the bowtie shape slot and the stepped square coaxial transition in a ridged waveguide cavity are utilised. The proposed antenna is well suited to use as an element for the phased array with the wide-angle scanning capability, and such array is a good candidate for the synthetic aperture radar (SAR) application. The development of an active solid phased array (ASPA) utilising bowtie slot waveguide antenna as the element is presented to illustrate the design procedure. The proposed ASPA is fabricated and the measured results are provided to demonstrate that the proposed array has good performance in field pattern and impedance matching over the full X-band (8-12 GHz). The ASPA's operating frequency band is cantered at 10 GHz with bandwidth 40%. This ASPA has been integrated into the latest SAR system with the maximum ground resolution of 0.05 m, and the SAR image is also provided at the end of this study.

Optimal trajectory selecting of receiver ship for shorebase-shipborne bistatic inverse synthetic aperture radar

As an important part of ocean surveillance systems, a typical application of inverse synthetic aperture radar (ISAR) images is target recognition and classification. Since the target projection on the ISAR image plane depends on relative radar–target geometry and the target’s motion, target recognition and classification becomes a more difficult task when the target is noncooperative. To overcome such a problem, shorebase-shipborne bistatic ISAR is proposed. Although the target of interest is noncooperative, the trajectory of the receiver ship platform can be controlled to change the relative radar–target geometry aiming to obtain a desired ISAR image projection. The Doppler frequency model of shorebase-shipborne bistatic ISAR imaging of the ship target in the presence of a low sea state is derived. An optimal trajectory of the receiver ship is presented that is able to obtain desired ISAR images, in which the cross-range coordinate of a scatterer is proportional to its coordinate x in space. The correctness of the proposed Doppler frequency model and effectiveness of trajectory selection are verified by means of several imaging simulations.

A Broadband Shared-Aperture L/S/X-Band Dual-Polarized Antenna for SAR Applications

This paper presents a novel tri-band antenna with dual polarizations and shared aperture. The proposed antenna operates at L-, S-, and X-bands with an approximate frequency ratio of 1:1.9:5.5. Three types of radiation elements resonating at different frequencies, including the square microstrip patch, microstrip dipole, and printed monopole, are interlaced in the same aperture. Printed monopole with a fork-shaped stub is adopted as the L-band element due to its wideband ability. Both the S- and X-band elements adopt a stacked structure to enhance the bandwidth. Besides, the S-band element employs a stepped gradient structure that can obtain a wider bandwidth than the normal microstrip dipole. The measured results agree well with the simulations. The measured impedance bandwidths of L-, S-, and X-bands are 21.7%, 11.6%, and 18.9%, respectively. The measured isolation at L-, S-, and X-bands is higher than 25, 38, and 29 dB, respectively. The antenna also exhibits an excellent radiation performance with the high cross-polarization discrimination at the three bands. To the best of the authors' knowledge, this is the first shared-aperture L-/S-/X-band dual-polarized broadband antenna in a true sense, which is useful for potential synthetic aperture radar applications.

Near-field 3D SAR Imaging Techniques Using a Scanning MIMO Array

Near-field 3-D imaging based on a scanning array is an important application of Synthetic Aperture Radar (SAR) 3-D imaging technology for civil use. Compared with Single-Input-Single-Output (SISO) arrays, a Multi-Input-Multi-Output (MIMO) scanning system is a special imaging method characterized by high imaging quality, high array efficiency, loose requirements for antenna spacing and low cost. In this paper, two imaging regimes, namely, MIMO-planar scanning and MIMO cylindrical scanning, are described in terms of signal models, imaging algorithms, experimental systems, and imaging results. The results show the great application potential of the imaging technology in various scenarios.

Comparative Analysis of Polarimetric SAR Calibration Methods

In the diverse applications of polarimetric Synthetic Aperture Radar (SAR) systems, it is a crucial to conduct polarimetric calibration, which aims to remove the radar system distortion effects prior to utilizing polarimetric SAR observations. The objective of this study is to evaluate the performance of different polarimetric calibration methods. Two widely used methods, the Van Zyl and Quegan methods, and one recently proposed method, such as the Villa method, have been selected among various calibration methods in literature. The selected methods have basic differences in their assumptions that are applied to the polarimetric system model. In order to evaluate the calibration performances under different system parameters and ground characteristics, comparative analysis of the calibration results were conducted on synthetic polarimetric SAR data and ALOS PALSAR quad-pol mode data. Based on the experimental results, the advantages and limitations of different methods were clarified, and a simple hybrid calibration method is presented to further improve the polarimetric calibration performance.

X-Band GaN High-Power Amplifier Using Hybrid Power Combining Technique for SAR Applications

An X-band high-power amplifier (HPA) based on gallium nitride (GaN) high electron mobility transistors (HEMTs) has been developed for synthetic aperture radar (SAR) applications. A hybrid power combining technique, including microstrip circuits and waveguides, is used to design the HPA. For reducing the size, four 50 W GaN HEMTs cascaded with one 1-to-4 power divider and one 4-to-1 power combiner form a 4-way power combined PCB circuits. For combing the high power and driving an antenna, two PCB circuits are combined by magic-T waveguides. The transmission efficiency of the power combining is approximately 80%. In the 10% duty cycle (pulse width 100 us), the output power of the HPA is over 200 W across the band of 9.5–9.8 GHz. The maximum output power is 230 W at 9.5 GHz, and the power gain is 8.3 dB at 46.1°C.

Sub-Nyquist SAR Based on Pseudo-Random Time-Space Modulation.

Sub-Nyquist sampling technology can ease the conflict between high resolution and wide swath in a synthetic aperture radar (SAR) system. However, the existing sub-Nyquist SAR imposes a constraint on the type of the observed scene and can only reconstruct the scene with small sparsity (i.e., number of significant coefficients). The information channel model of microwave imaging radar based on information theory, in which scene, echo, and the mapping relation between the two correspond to information source, sink, and channel, is built, and noisy-channel coding theorem explains the reason for the aforementioned under this model. To allow the wider application of sub-Nyquist SAR, this paper proposes sub-Nyquist SAR based on pseudo-random space-time modulation. This modulation is the spatial and temporal phase modulation to the traditional SAR raw data and can increase the mutual information of information source and sink so that the scenes with large sparsity can be reconstructed. Simulations of scenes with different sparsity, e.g., an ocean with several ships and urban scenes, were run to verify the validity of our proposed method, and the results show that the scenes with large sparsity can be successfully reconstructed.

Using TanDEM-X Pursuit Monostatic Observations with a Large Perpendicular Baseline to Extract Glacial Topography

Space-based Interferometric Synthetic Aperture Radar (InSAR) applications have been widely used to monitor the cryosphere over past decades. Owing to temporal decorrelation, interferometric coherence often severely degrades on fast moving glaciers. TanDEM-X observations can overcome the temporal decorrelation because of their simultaneous measurements by two satellite constellations. In this study, we used the TanDEM-X pursuit monostatic mode with large baseline formation following a scientific phase timeline to develop highly precise topographic elevation models of the Petermann Glacier of Northwest Greenland. The large baseline provided the advantage of extracting topographic information over low relief areas, such as the surface of a glacier. As expected, coherent interferometric phases (>0.8) were well maintained over the glaciers, despite their fast movement, due to the nearly simultaneous TanDEM-X measurements. The height ambiguity, which was defined as the altitude difference corresponding to a 2π phase change of the flattened interferogram, of the dataset was 10.63 m, which was favorable for extracting topography in a low relief region. We validated the TanDEM-X derived glacial topography by comparing it to the SAR/Interferometric radar altimeter observations acquired by CryoSat-2 and the IceBridge Airborne Topographic Mapper laser altimeter measurements. Both observations showed very good correlation within a few meters of the offsets (−12.5~−3.1 m), with respect to the derived glacial topography. Routine TanDEM-X observations will be very useful to better understand the dynamics of glacial movements and topographic change.

The Dependence of Sea SAR Image Distribution Parameters on Surface Wave Characteristics

Modeling the statistical distribution of synthetic aperture radar (SAR) images is essential for sea target detection, which is an important aspect of marine SAR applications. The main goal of this study is to determine the effects of sea states and surface wave texture characteristics on the statistical distributions of sea SAR images. A statistical analysis of the Envisat Advanced Synthetic Aperture Radar (ASAR) wave mode images (imagettes), covering a variety of sea states and wave conditions, was carried out to investigate the suitability of the statistical distributions often used in the literature for sea states parameters. The results revealed the variation in the distribution parameters in terms of their azimuthal cutoff wavelength (ACW) and the peak-to-background ratio (PBR) of the SAR image intensity spectra. The shape parameters of Gamma and Weibull distribution are sensitive and monotonously decreasing with respect to PBR, while the scale parameter is sensitive to ACW. The K distribution was shown to perform well, with both high and stable accuracy. The results of this paper provide a parameterized scheme for sea state classifications and can potentially be used for choosing the most suitable distribution model according to sea state when performing sea target detection.

Evaluation of Chinese Quad-polarization Gaofen-3 SAR Wave Mode Data for Significant Wave Height Retrieval

Our work describes the accuracy of Chinese quad-polarization Gaofen-3 (GF-3) synthetic aperture radar (SAR) wave mode data for wave retrieval and provides guidance for the operational applications of GF-3 SAR. In this study, we evaluated the accuracy of the SAR-derived significant wave height (SWH) from 10,514 GF-3 SAR images with visible wave streaks acquired in wave mode by using the existing wave retrieval algorithms, e.g., the theoretical-based algorithm parameterized first-guess spectrum method (PFSM), the empirical algorithm CSAR_WAVE2 for VV-polarization, and the algorithm for quad-polarization (Q-P). The retrieved SWHs were compared with the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis field with 0.125° grids. The root mean square error (RMSE) of the SWH is 0.57 m, found using CSAR_WAVE2, and this RMSE value was less than the RMSE values for the analysis results achieved with the PFSM and Q-P algorithms. The statistical analysis also indicated that wind speed had little impact on the bias with increasing wind speed. However, the retrieval tended to overestimate when the SWH was smaller than 2.5 m and underestimate with an increasing SWH. This behavior provides a perspective of the improvement needed for the SWH retrieval algorithm using the GF-3 SAR acquired in wave mode.

Change Detection of Polarimetric SAR Images Based on the Integration of Improved Watershed and MRF Segmentation Approaches

Land cover and land use change detection is one of the most important applications of polarimetric synthetic aperture radar (PolSAR) data. In a pixel-based method, the speckle effects and the lack of consideration for spatial information produce false alarms. Therefore, an object-based image analysis has been used to overcome these weaknesses. This paper proposes a novel hybrid image segmentation algorithm for improving the accuracy of land cover change detection. This method consists of three steps. First, the segmentation of two PolSAR images by integrating segmentation techniques, namely, region-based improved watershed and an edge-based coupled Markov random field (MRF). Second, the selection of the optimal ratios of polarimetric features based on a genetic algorithm and Jefferies–Matusita distance criteria. Third, the binary classification of image objects using the ratio of mean pixel values of the corresponding image objects. Compared with the conventional watershed and multiresolution segmentation methods, the improved watershed reduces the speckle effects in PolSAR images and avoids the oversegmentation problem. The use of coupled MRF detects the most accurate edge positions and improves the hybrid segmentation approach. The proposed change detection method, as applied on unmanned aerial vehicle synthetic aperture full polarimetric images, demonstrates an overall accuracy of 90.14% and a kappa coefficient of 0.80. Compared with the pixel-based change detection method based on a support vector machine classifier for the selected subspace of ratios of polarimetric features, the proposed method improved the overall accuracy by 8.72% and the kappa coefficient by 0.18.

Satellite Imagery for Classification of Rice Growth Phase Using Freeman Decomposition in Indramayu, West Java, Indonesia

Monitoring at every growth of rice plants is an important information for determining the grain pro-duction estimation of rice. Monitoring must to be have timely work on the rice plant development. However, timely monitoring and the high accuracy of information is a challenge in remote sensing based on rice agriculture monitoring and observation. With increased quality of synthetic aperture radar (SAR) systems utilizing polarimetric information recently, the development and applications of polarimetric SAR (PolSAR) are one of the current major topics in radar remote sensing. The ad-vantages provided by PolSAR data for agricultural monitoring have been extensively studied for applications such as crop type classification and mapping, crop phenology monitoring, productivity assessment based on the sensitivity of polarimetric parameters to indicators of crop conditions. Freeman and Durden successfully decomposed fully PolSAR data into three components: Single bounce, double bounce, and volume scattering. The three-component scattering provide features for distinguishing between different surface cover types. These sensitivities assist in the identification of growing phase. The observed growing phase development in time series, reflected in the consistent temporal trends in scattering, was generally in agreement with crop phenological development stages. Supervised classification was performed on repeat-pass Radarsat-2 images, with an overall classification accuracy of 77.27% achieved using time series Fine beam data. The study demonstrated that Radarsat-2 Fine mode data provide useful information for crop monitoring and classification of rice plants.

Synthetic Aperture Radar Processing Approach for Simultaneous Target Detection and Image Formation.

Finding out interested targets from synthetic aperture radar (SAR) imagery is an attractive but challenging problem in SAR application. Traditional target detection is independent on SAR imaging process, which is purposeless and unnecessary. Hence, a new SAR processing approach for simultaneous target detection and image formation is proposed in this paper. This approach is based on SAR imagery formation in time domain and human visual saliency detection. First, a series of sub-aperture SAR images with resolutions from low to high are generated by the time domain SAR imaging method. Then, those multiresolution SAR images are detected by the visual saliency processing, and the corresponding intermediate saliency maps are obtained. The saliency maps are accumulated until the result with a sufficient confidence level. After some screening operations, the target regions on the imaging scene are located, and only these regions are focused with full aperture integration. Finally, we can get the SAR imagery with high-resolution detected target regions but low-resolution clutter background. Experimental results have shown the superiority of the proposed approach for simultaneous target detection and image formation.

Crop Classification Based on Differential Characteristics of <inline-formula> <tex-math notation="LaTeX">$H/\alpha$ </tex-math> </inline-formula> Scattering Parameters for Multitemporal Quad- and Dual-Polarization SAR Images

Crop-type classification is one of the most significant applications in polarimetric synthetic aperture radar (PolSAR) imagery. As a remote sensing technique, PolSAR has been proved to have the ability to provide high-resolution information of illustrated objects. However, single-temporal PolSAR data are restricted to provide sufficient information for crop identification due to the complicated condition of varying morphology within various growing stages. With an increasing number of spaceborne PolSAR systems launched, a large amount of real PolSAR data are being generated and used to provide great opportunities for multitemporal analysis. The main contribution of this paper is to improve crop classification accuracy with various features of classical $H$ / $\alpha$ parameters. First, in order to deal with dual-PolSAR data, $H$ / $\alpha$ decomposition algorithm for quad-PolSAR is modified to suit to the case of dual polarization. Second, according to the differential scattering characteristics of main crops, a new parameter is innovatively defined to measure the differential characteristics in the $H$ / $\alpha$ classification plane. Third, crop types are discriminated by applying a supervised classification method with the newly defined parameter. Furthermore, the correctness of the parameter is verified with simulated and real Sentinel-1 data as well as AirSAR data. Finally, the performances of the classification method are investigated by the comparison with complex Wishart, Freeman–Wishart, and support vector machine (SVM) classifiers. Hence, the experimental results show that the proposed method and SVM classifier with the newly defined parameter have the ability to improve crop classification accuracy.

A Concurrent and Hierarchy Target Learning Architecture for Classification in SAR Application.

This article discusses the issue of Automatic Target Recognition (ATR) on Synthetic Aperture Radar (SAR) images. Through learning the hierarchy of features automatically from a massive amount of training data, learning networks such as Convolutional Neural Networks (CNN) has recently achieved state-of-the-art results in many tasks. To extract better features about SAR targets, and to obtain better accuracies, a new framework is proposed: First, three CNN models based on different convolution and pooling kernel sizes are proposed. Second, they are applied simultaneously on the SAR images to generate image features via extracting CNN features from different layers in two scenarios. In the first scenario, the activation vectors obtained from fully connected layers are considered as the final image features; in the second scenario, dense features are extracted from the last convolutional layer and then encoded into global image features through one of the commonly used feature coding approaches, which is Fisher Vectors (FVs). Finally, different combination and fusion approaches between the two sets of experiments are considered to construct the final representation of the SAR images for final classification. Extensive experiments on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset are conducted. Experimental results prove the capability of the proposed method, as compared to several state-of-the-art methods.