Synthetic Aperture Radar Image Generation Research Articles

High-Precision GPU-Accelerated Simulation Algorithm for Targets under Non-Uniform Cluttered Backgrounds

This article presents a high-precision airborne video synthetic aperture radar (SAR) raw echo simulation method aimed at addressing the issue of simulation accuracy in video SAR image generation. The proposed method employs separate techniques for simulating targets and ground clutter, utilizing pre-existing SAR images for clutter simulation and employing the shooting and bouncing rays (SBR) approach to generate target echoes. Additionally, the method accounts for target-generated shadows to enhance the realism of the simulation results. The fast simulation algorithm is implemented using the C++ programming language and the Accelerated Massive Parallelism (AMP) framework, providing a fusion technique for integrating clutter and target simulations. By combining the two types of simulated data to form the final SAR image, the method achieves efficient and accurate simulation technology. Experimental results demonstrate that this method not only improves computational speed but also ensures the accuracy and stability of the simulation outcomes. This research holds significant implications for the development of algorithms pertaining to video SAR target detection and tracking, providing robust support for practical applications.

Ship Detection in Low-Quality SAR Images via an Unsupervised Domain Adaption Method

Ship detection in low-quality Synthetic Aperture Radar (SAR) images poses a persistent challenge. Noise signals in complex environments disrupt imaging conditions, hindering SAR systems from acquiring precise target information, thereby significantly compromising the performance of detectors. Some methods mitigate interference via denoising techniques, while others introduce noise using classical methods to learn target features in the presence of noise. This conundrum is regarded as a cross-domain problem in this paper; a ship detection method in low-quality images is proposed to learn features of targets and shrink serious deterioration of detection performance by utilizing Generative Adversarial Networks (GANs). First, an image-to-image translation task is implemented using CycleGAN to generate low-quality SAR images with complex interference from the source domain to the target domain. Second, with the annotation inheritance, these generated SAR images participate in a training process to improve the detection accuracy and model robustness. Multiple experiments indicate that the proposed method conspicuously improves the detection performance and efficaciously reduces the missed detection rate in the SAR ship detection task. This cross-domain approach achieved outstanding improvements in the form of 11.0% mAP and 3.22% mAP on the GaoFen-3 ship dataset and SRSSD-V1.0, respectively. In addition, the characteristics and potentials of near-shore and off-shore SAR image reconstruction with style transfer based on Generative Adversarial Networks were explored and analyzed in this work.

Curved Synthetic Aperture Radar for Near-Field Terahertz Imaging

High resolution imaging technique will play an essential role in the future smart imaging applications to provide potential solutions for accurate detection, monitoring and classification of the target. The potential of synthetic aperture radar (SAR) systems to acquire high resolution images has opened new frontiers and led to its exploration in new applications. The conventional microwave radar imaging suffers limited range resolution in the sub-millimeter wave range and optical imaging methods are constrained by the diffraction limit. In the terahertz (THz) regime of the spectrum, the SAR systems could be operated in all the weather, all the time and are capable of penetrating through clouds, smoke, dust etc. to achieve high resolution images beyond the diffraction limit. The curved SAR presents a SAR system created by circularly encircling the target scene. The Circular SAR and Spiral SAR which are based on a circular and cylindrical spiral scanning trajectory, respectively are the scanning methods used for curved SAR. In this paper, we have explored a new toolbox that enables the rapid development of near-field THz-SAR imaging systems and generation of large near-field THz imaging scenarios with the goal to support and facilitate data driven research such as deep learning in the THz community through synthetically generated SAR images as well as its potential for THz based high resolution and 2D/3D near-field imaging application. Initially, the potential advantages of THz-SAR over Microwave SAR and optical imaging methods are presented. Further, a frequency modulated continuous wave (FMCW) radar testbed is simulated in multi-input multi-output (MIMO) configuration for the target frequency band. Moreover, the circular and cylindrical SAR scanning methods are explored for near-field imaging of point targets to obtain 2D and 3D THz-SAR images, respectively. Also, the radar images are reconstructed using the Back Propagation and Polar Formatting Algorithms, focusing on short range applications like indoor environments. Finally, the SAR performance evaluation metrics are reported and a roadmap for future work is presented.

Two-Way Generation of High-Resolution EO and SAR Images via Dual Distortion-Adaptive GANs

Synthetic aperture radar (SAR) provides an all-weather and all-time imaging platform, which is more reliable than electro-optical (EO) remote sensing imagery under extreme weather/lighting conditions. While many large-scale EO-based remote sensing datasets have been released for computer vision tasks, there are few publicly available SAR image datasets due to the high costs associated with acquisition and labeling. Recent works have applied deep learning methods for image translation between SAR and EO. However, the effectiveness of those techniques on high-resolution images has been hindered by a common limitation. Non-linear geometric distortions, induced by different imaging principles of optical and radar sensors, have caused insufficient pixel-wise correspondence between an EO-SAR patch pair. Such a phenomenon is not prominent in low-resolution EO-SAR datasets, e.g., SEN1-2, one of the most frequently used datasets, and thus has been seldom discussed. To address this issue, a new dataset SN6-SAROPT with sub-meter resolution is introduced, and a novel image translation algorithm designed to tackle geometric distortions adaptively is proposed in this paper. Extensive experiments have been conducted to evaluate the proposed algorithm, and the results have validated its superiority over other methods for both SAR to EO (S2E) and EO to SAR (E2S) tasks, especially for urban areas in high-resolution images.

SAR image formation scheme implementation and endorsement sprouting from Level-0 data decoding

Synthetic Aperture Radar (SAR) data processing evolving from level-0 raw data is complicated, especially in data decoding, manifesting in obtaining a well-focused SAR image. This paper is intended to present a complete MATLAB-based SAR data processing tool, which helps the end-user to treat simply the steps of image generation. This paper would enrich the research community in the field of SAR processors, especially in the area of understanding, handling, and developing a SAR processor based on space packet protocol standard (STD 01) used in many SAR systems such as Sentinel-1, ERS-1, CubeL, JPSS-2, 3, and 4. Also, this work opens the door for researchers to decode other space packet protocol standards and even to create an algorithm based on fully understanding the image formation algorithm from its roots. Moreover, the work in this paper could be a stepping-stone for the beginner in the field of SAR signal processing to become familiarized with SAR image generation procedures. The level-0 raw data used in this paper is based on Sentinel 1 SAR satellite obtained from the European Space Agency Copernicus website, a free open-source for level-0 and level-1 data types. The MATLAB program allows users to compare their generated image with the level-1 single-look complex (S1-L1-SLC) image utilizing entropy, contrast, and sharpness image quality metrics.The results showed that the images produced by the proposed algorithm are comparable to Sentinel-1 level-1 SAR images for the same scene and achieved satisfactory accuracy under the requirements for image quality.

Conversion from multi-spectral data into SAR data with Deep Convolution Neural Architecture using Generative Adversarial Network

Synthetic Aperture Radar (SAR) is a high-resolution radar extensively used for aerial and terrestrial images in numerous applications. The SAR images are all weather and day/night capable. It is extensively used in defense/military systems for surveillance, target detection and target tracking. In current times, the need for deep learning algorithms has increased due to its capability of generating SAR images, as it is very intensive and expensive task for the acquisition and formation of SAR images through sensors. Thus, alternative methods are required to produce SAR image data. In this paper, generative adversarial networks (GAN) are used to generate SAR images from a limited pre-existing dataset. This paper presents a novel way of translating optical images to SAR images, for the generation of paired, unpaired data set. Due to the unavailability of an open-source paired, unpaired image dataset, a synthetic paired image dataset is created for proving the concepts of CycleGAN and Pix-to-Pix algorithm by fine tuning the algorithms using generated datasets. The paper tried tackling this problem in two ways, both paired and unpaired image-to-image translation.

Reference Map Generation for Synthetic Aperture Radar Based on CycleGAN

The real-time closed-loop simulator of Synthetic Aperture Radar (SAR) needs to input the SAR radar reference map as the simulation target scene information in the process of simulation. However, it is difficult to obtain the scene map of SAR radar in the detection range that is realistic and consistent with the task target, and the data sets are limited. In this article, Cycle-Consistent Adversarial Networks (CycleGAN) is applied to realize the migration learning of SAR radar reference map and solve the task required of a directional generation of SAR images.

Multi-Level Alignment Network for Cross-Domain Ship Detection

Ship detection is an important research topic in the field of remote sensing. Compared with optical detection methods, Synthetic Aperture Radar (SAR) ship detection can penetrate clouds to detect hidden ships in all-day and all-weather. Currently, the state-of-the-art methods exploit convolutional neural networks to train ship detectors, which require a considerable labeled dataset. However, it is difficult to label the SAR images because of expensive labor and well-trained experts. To address the above limitations, this paper explores a cross-domain ship detection task, which adapts the detector from labeled optical images to unlabeled SAR images. There is a significant visual difference between SAR images and optical images. To achieve cross-domain detection, the multi-level alignment network, which includes image-level, convolution-level, and instance-level, is proposed to reduce the large domain shift. First, image-level alignment exploits generative adversarial networks to generate SAR images from the optical images. Then, the generated SAR images and the real SAR images are used to train the detector. To further minimize domain distribution shift, the detector integrates convolution-level alignment and instance-level alignment. Convolution-level alignment trains the domain classifier on each activation of the convolutional features, which minimizes the domain distance to learn domain-invariant features. Instance-level alignment reduces domain distribution shift on the features extracted from the region proposals. The entire multi-level alignment network is trained end-to-end and its effectiveness is proved on multiple cross-domain ship detection datasets.

LIME-Based Data Selection Method for SAR Images Generation Using GAN

Deep learning has obtained remarkable achievements in computer vision, especially image and video processing. However, in synthetic aperture radar (SAR) image recognition, the application of DNNs is usually restricted due to data insufficiency. To augment datasets, generative adversarial networks (GANs) are usually used to generate numerous photo-realistic SAR images. Although there are many pixel-level metrics to measure GAN’s performance from the quality of generated SAR images, there are few measurements to evaluate whether the generated SAR images include the most representative features of the target. In this case, the classifier probably categorizes a SAR image into the corresponding class based on “wrong” criterion, i.e., “Clever Hans”. In this paper, local interpretable model-agnostic explanation (LIME) is innovatively utilized to evaluate whether a generated SAR image possessed the most representative features of a specific kind of target. Firstly, LIME is used to visualize positive contributions of the input SAR image to the correct prediction of the classifier. Subsequently, these representative SAR images can be selected handily by evaluating how much the positive contribution region matches the target. Experimental results demonstrate that the proposed method can ally “Clever Hans” phenomenon greatly caused by the spurious relationship between generated SAR images and the corresponding classes.

Learning to Generate SAR Images With Adversarial Autoencoder

Deep learning-based synthetic aperture radar (SAR) target recognition often suffers from sparsely distributed training samples and rapid angular variations due to scattering scintillation. Thus, data-driven SAR target recognition is considered a typical few-shot learning (FSL) task. This article first reviews the key issues of FSL and provides a definition of the FSL task. A novel adversarial autoencoder (AAE) is then proposed as an SAR representation and generation network. It consists of a generator network that decodes target knowledge to SAR images and an adversarial discriminator network that not only learns to discriminate “fake” generated images from real ones but also encodes the input SAR image back to target knowledge. The discriminator employs progressively expanding convolution layers and a corresponding layer-by-layer training strategy. It uses two cyclic loss functions to enforce consistency between the inputs and outputs. Moreover, rotated cropping is introduced as a mechanism to address the challenge of representing the target orientation. The moving and stationary Target recognition (MSTAR) 7-target dataset is used to evaluate the AAE’s performance, and the results demonstrate its ability to generate SAR images with aspect angular diversity. Using only 90 training samples with at least 25° of orientation interval, the trained AAE is able to generate the remaining 1748 samples of other orientation angles with an unprecedented level of fidelity. Thus, it can be used for data augmentation in SAR target recognition FSL tasks. Our experimental results show that the AAE could boost the test accuracy by 5.77%.

PeaceGAN: A GAN-Based Multi-Task Learning Method for SAR Target Image Generation with a Pose Estimator and an Auxiliary Classifier

Although generative adversarial networks (GANs) are successfully applied to diverse fields, training GANs on synthetic aperture radar (SAR) data is a challenging task due to speckle noise. On the one hand, in a learning perspective of human perception, it is natural to learn a task by using information from multiple sources. However, in the previous GAN works on SAR image generation, information on target classes has only been used. Due to the backscattering characteristics of SAR signals, the structures of SAR images are strongly dependent on their pose angles. Nevertheless, the pose angle information has not been incorporated into GAN models for SAR images. In this paper, we propose a novel GAN-based multi-task learning (MTL) method for SAR target image generation, called PeaceGAN, that has two additional structures, a pose estimator and an auxiliary classifier, at the side of its discriminator in order to effectively combine the pose and class information via MTL. Extensive experiments showed that the proposed MTL framework can help the PeaceGAN’s generator effectively learn the distributions of SAR images so that it can better generate the SAR target images more faithfully at intended pose angles for desired target classes in comparison with the recent state-of-the-art methods.

Small UAV-based SAR system using low-cost radar, position, and attitude sensors with onboard imaging capability

AbstractThis paper describes a small unmanned aerial vehicle (UAV)-based synthetic aperture radar (SAR) system using low-cost radar (5–6 GHz), position (GNSS/RTK) and attitude (IMU) sensors for the generation of high-resolution images. Measurements using straight as well as highly curved flight trajectories and varying flight speeds are presented, showing range and cross-range lobe-widths close to the theoretical limits. An analysis of the improvements obtained by the use of attitude angles (roll, pitch, and yaw), to correct for the relative offsets in antenna positions as the UAV moves, is included. A capability to generate SAR images onboard with the back-projection algorithm has been implemented using a GPU accelerated single-board computer. Generated images are transmitted to ground using a Wi-Fi data link.

Multiplicative-additive despeckling in SAR images

Visual and automatic analyses using synthetic aperture radar (SAR) images are challenging because of inherently formed speckle noise. Thus, reducing speckle noise in SAR images is an important research area for SAR image analysis. During speckle noise reduction, homogeneous regions should be smoothed while details such as edges and point scatterers need to be preserved. General speckle noise model contains gamma distributed multiplicative part which is dominant and Gaussian distributed additive part which is in low amount and mostly neglected in literature. In this study, a novel sparsity-driven speckle reduction method is proposed that takes both multiplicative noise model and additive noise model into consideration. The proposed speckle reduction method uses a cost function with multiplicative and additive data terms besides the total variation smoothness term. Also, an efficient and stable numerical minimization scheme is proposed for the proposed cost function that deals with multiplicative and additive noise. Speckle reduction performance of the proposed method is shown on synthetically generated SAR images and real-world SAR images.

Synthetic Aperture Radar Image Generation With Deep Generative Models

A variety of machine learning approaches have been applied to synthetic aperture radar (SAR) automatic target recognition. The performances of these approaches rely strongly on the quality and quantity of training data. In real-world applications, however, it is challenging to obtain sufficient data suitable for these approaches. To alleviate this problem, a novel deep generative model for SAR image generation is proposed, which is an extension of Wasserstein autoencoder. The network structure and reconstruction loss function of the model have been improved according to the characteristics of SAR images. The experimental results demonstrate that our model is superior to other classical generative models in SAR image generation. The generated images can be directly used as training samples, thereby extending the training data set and improving the recognition accuracy.

Speckle-Noise-Invariant Convolutional Neural Network for SAR Target Recognition

Speckle noise is inherent to synthetic aperture radar (SAR) images and degrades the target recognition performance. Deep learning based on convolutional neural networks (CNNs) has been widely applied for SAR target recognition, but the extracted features are still sensitive to speckle noise. In addition, speckle noise has been seldom considered in such CNN-based approaches. In this letter, we propose a speckle-noise-invariant CNN that employs regularization for minimizing feature variations caused by this noise. Before CNN training, we performed SAR image despeckling using the improved Lee sigma filter for feature extraction. Then, we generated SAR images for CNN training by adding speckle noise to the despeckled images. The proposed regularization improves both the feature robustness to speckle noise and SAR target recognition. Experiments on the moving and stationary target acquisition and recognition database demonstrate that the proposed CNN notably improves the classification accuracy compared with the conventional methods.

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.

Moisture determination of concrete panel using SAR imaging and the K-R-I transform

Moisture content inside concrete affects the properties and behaviors of Portland cement concrete. It also indicates the likelihood of structural damages (e.g., steel corrosion) in reinforced and prestressed concrete structures. While several laboratory techniques are available for moisture determination, it is a challenging task to estimate the moisture content of concrete in the field without using embedded moisture sensors. In this paper, synthetic aperture radar (SAR) imaging and the K-R-I (curvature-area-amplitude) transform were applied to a concrete panel specimen (water-to-cement ratio = 0.45) for moisture determination. A 10.5 GHz center frequency radar system was used to generate SAR images of the concrete panel at various moisture levels from 0% to 3.85% (by mass). Quantitative analysis of SAR images was carried out by the K-R-I transform to understand the simultaneous change of SAR amplitude and shape at different moisture levels. It was found that integrated SAR amplitude and average maximum SAR amplitude both increase nonlinearly with the increase of moisture content in the concrete panel. Spatial distribution of SAR amplitudes can be used to indicate subsurface moisture distribution in concrete. The area-amplitude (R-I) curve of SAR images quantifies the relationship between moisture content and its distribution.

Acceleration of monostatic shooting and bouncing ray method for inverse synthetic aperture radar image generation

ABSTRACTThe enhanced shooting and bouncing ray (SBR) technique based on the bistatic approximation, which is called the one‐shot inverse synthetic aperture radar (ISAR) technique, has been devised to accelerate the conventional monostatic SBR (CM‐SBR) technique, which is suitable for inverse synthetic aperture radar (ISAR) imaging. However, because of the bistatic approximation, this method results in the distortion of ISAR images. Therefore, to obtain an accurate image, radar returns should be obtained from the monostatic configuration. This paper describes a new method, the ray‐travel map (RTM), which accelerates the CM‐SBR technique while simultaneously considering both the CM‐SBR and the ISAR configuration. Then, we demonstrate that the proposed method can significantly enhance the speed of CM‐SBR, while providing accurate distortion‐free ISAR image. © 2016 Wiley Periodicals, Inc. Microwave Opt Technol Lett 58:1848–1855, 2016

High-Resolution SAR Image Generation by Subaperture Processing of FMCW Radar Signal

In this letter, we present a subaperture-based processing scheme for synthetic aperture radar (SAR) imaging using frequency-modulated continuous-wave (FMCW) radar transmission. An efficient two-stage azimuth subaperture processing scheme is proposed for high-resolution SAR image generation. A new bulk range-cell-migration-correction technique is proposed. The algorithms are validated with point target simulations at the X-band and also using real data samples at the C-band. The results show that subaperture-based processing can be used to image extended ground scenes without mosaicing.

STK를 이용한 고해상도 위성 SAR 운용 및 표적물 추출 기법

A comprehensive SAR(Synthetic Aperture Radar) simulation is considered to be a complicated task since a full knowledge of the signal propagation characteristics, antenna pattern, system internal errors and interference noises should be taken into account. In high resolution target application modes, the time varying nature of target RCS(Radar Cross Section) strongly affects the generated SAR images. In this paper, in-depth SAR simulations are performed and analyzed incorporating the STK tools and MATLAB software. STK provides realistic orbit parameters while its radar module helps to extract accurate radiometric parameters of ground targets. SAR raw data corresponding to a given target is generated and processed using MATLAB simulator. The performance is measured by PSLR(Peak Sidelobe Ratio) and ISLR(Integrated Sidelobe Ratio) for a point target, which can be used as reference parameters for accurate radiometric calibration. Finally, high resolution target simulations are performed by adopting time varying target RCS characteristics.