Tomographic synthetic aperture radar (TomoSAR) shows great potential for high-quality 3-D mapping, especially in urban areas. As TomoSAR reconstruction methods advance into the deep learning (DL) era, current studies have demonstrated DL’s strengths in both precision and efficiency. However, for reconstructing urban areas with prominent spatial features from building structures, current studies focus on pixel-by-pixel reconstruction without leveraging the potential benefits of these features. In this context, an exploratory study to introduce spatial feature regularization in DL reconstruction is proposed for the first time, focusing on feature description, modeling, and regularization. Spatial features are analyzed and summarized by sharp edges and regular geometric shapes within the scene. To model these features, 2-D slices are used as the basic reconstruction units, and a general intraslice and interslice strategy is proposed to harness features within and between slices. Two-dimensional slices are fused into the entire 3-D scene. Two methods of fusion are designed: parallel and serial. To regularize these features, a new computational framework called light reconstruction and enhancement is designed, which includes two stages: light reconstruction with sparsity feature regularization and enhancement with spatial feature regularization. Finally, to evaluate performance, we design an extensive evaluation framework. A newly self-constructed compound urban building simulation dataset, combined with two public measured data, forms six different tests ranging from a classical close point resolution test to a diverse urban landscape challenge test. Evaluation results reveal the effectiveness of the designs and the boost provided by spatial feature regularization, resulting in higher reconstruction precision, more complete building spatial structure retrieval, and fewer outliers.

Tomographic synthetic aperture radar (tomoSAR) imaging is a powerful tool for urban 3D reconstruction. While recent deep learning methods have improved reconstruction quality, their reliance on simulated measurement-scene-image pairs for supervised training raises concerns over robustness in real-world scenarios due to the distribution shifts, such as varying observation geometry, observed scene distributions, and signal/noise levels. These concerns have received limited attention until now, motivating us to explore an alternative approach that leverages the strength of deep learning for tomoSAR reconstruction without requiring paired measurement and scene-image training data. Therefore, we propose the Unified Learning and Reconstruction (ULAR), a model-driven method for robust tomoSAR reconstruction, trained without such paired data. ULAR integrates physical-model consistency with scene feature regularization (capturing spatial structures) in a unified optimization process. Specifically, it jointly performs image reconstruction and spatial structure refinement by alternating between physics-guided updates and mainly self-supervised spatial-structure learning. The approach incorporates two complementary components for spatial-structure learning: a one-step self-supervised generative model for local spatial structures and a pretrained denoiser for nonlocal ones. And the denoiser is further enhanced with equivariance properties to improve its robustness. Experimental results on both simulated and real measured datasets demonstrate that ULAR achieves reconstruction accuracy comparable to supervised methods, and even surpasses them when distribution shifts exist, revealing strong robustness while not relying on simulated measurement-ground truth paired data. These results demonstrate the robustness of self-supervised learning for tomoSAR reconstruction, while highlighting its potential for better practicality and reliability in real-world applications.

Millimeter-wave (mmW) automotive radar imaging technology shows significant promise in advanced driver assistance systems (ADAS). Super-resolution imaging methods can be employed the limited aperture length of automotive radar to improve azimuth (angular) resolution. However, automotive radar images typically exhibit large dynamic range (LDR) and large scene (LS), leading to pay extensive computational complexity and storage demands when striving for higher image quality. To tackle this challenge, a fast$l_{q}$sparsity-driven imaging method with adaptive-focusing framework (FLSD-AF) for mmWave automotive radar super-resolution imaging in this article. First, in AF framework, a detect-before-imaging (DBI) is proposed to make echo data to adaptive focused on potential target area (PTR), thereby reducing the dimension of the effective data to reduce computational complexity and storage demands. Second, a subspace-phase-compensation (SPC) is proposed to reduces storage demands of the measurement matrix by addressing the imaging model mismatch in near-field under LS. Finally, a fast$l_{q}$sparsity-driven (FLSD) imaging method is proposed. It employs$l_{q}$-norm nonconvex penalty function to address the biased problem to improve imaging quality under LDR, meanwhile the computational complexity of the matrix operation is greatly reduced by utilizing joint Kailath-Variant (K-V) formula and Gohberg-Semencul (G-S) factorization. In summary, the proposed FLSD-AF not only substantially enhances the imaging performance, but also significantly diminishes the storage demands and computational complexity under LDR and LS. The results of simulations and experimental data all verify the proposed method.

The near-field array synthetic aperture radar (SAR) imaging method that’s based on deep neural networks has significantly advanced imaging accuracy and efficiency compared to traditional techniques like matched filtering and sparse reconstruction. However, it currently relies on supervised learning, which is affected by differences between simulated and real data. To address the issue, we introduces an unsupervised approach based on generative models of latent variables for near-field array SAR imaging. By focusing on generating target image distributions, this method bypasses the need for simulated training data. Instead, it leverages the concept of generative models with latent variables, using a prior auxiliary variable to construct a decoding neural network that transforms these latent variables into target images. In addition, a model-driven loss function is designed based on physical priors related to the linear correspondence between echoes and target images in the SAR measurement process.To enhance image quality further, sparse constraints (L1) loss function is integrated into the approach’s final loss function. Experimental validation using actual millimeter-wave near-field array SAR data demonstrates the effectiveness of this unsupervised imaging method. It offers advantages such as not relying on simulated data for training, suitability for diverse target types, superior imaging accuracy compared to traditional methods, and the ability to maintain high accuracy even at low sampling rates (10%).

Array synthetic aperture radar (SAR) three-dimensional (3D) imaging can obtain 3D information of the target region, which is widely used in environmental monitoring and scattering information measurement. In recent years, with the development of compressed sensing (CS) theory, sparse signal processing is used in array SAR 3D imaging. Compared with matched filter (MF), sparse SAR imaging can effectively improve image quality. However, sparse imaging based on handcrafted regularization functions suffers from target information loss in few observed SAR data. Therefore, in this article, a general 3D sparse imaging framework based on Regulation by Denoising (RED) and proximal gradient descent type method for array SAR is presented. Firstly, we construct explicit prior terms via state-of-the-art denoising operators instead of regularization functions, which can improve the accuracy of sparse reconstruction and preserve the structure information of the target. Then, different proximal gradient descent type methods are presented, including a generalized alternating projection (GAP) and an alternating direction method of multiplier (ADMM), which is suitable for high-dimensional data processing. Additionally, the proposed method has robust convergence, which can achieve sparse reconstruction of 3D SAR in few observed SAR data. Extensive simulations and real data experiments are conducted to analyze the performance of the proposed method. The experimental results show that the proposed method has superior sparse reconstruction performance.

Millimeter-wave (mmW) automotive synthetic aperture radar (Auto-SAR) technology holds significant promise for advanced driver assistance systems (ADASs). Sparse imaging methods can improve the quality of Auto-SAR images, such as suppressing sidelobes and noise. However, the$l_{1}$convex regularization-based sparse imaging methods suffer from the bias estimation, which reduces the target amplitude and ignores the association between scatterers, weakening the target structure. To address these issues, we proposed joint generalized$l_{q}$and convolutional (Glq-Con) regularization to enhance mmW Auto-SAR sparse imaging in this article. First, to improve the target amplitude, we propose utilizing the nonconvexity of Glq to reduce the bias effect; meanwhile, the global convergence of Glq ensures the imaging accuracy. Then, considering the continuity of the imaging target in driving scenes, we propose to utilize convolution regularization to modify the previously reconstructed amplitude of Glq to improve the target structure. Besides, to reduce computational complexity, we establish an efficient sparse imaging model. In this model, the fast Fourier transform (FFT) operator is employed to approximate complex matrix operation in the iterative process. We also use an efficient optimizer to solve the imaging model. Finally, both simulations and measured typical driving scenario experiments demonstrate that the proposed method significantly enhanced the Auto-SAR image, especially for the targets of weak scatterers.

TomoSAR imaging is a classical inverse problem. Learning to optimize is a newly emerging technique for solving such problems in a deep-learning way. This technique may facilitate the efficiency and accuracy of the inverse process. In this research, we apply this framework to TomoSAR imaging and aim to enhance spatial structures within the framework. We establish a two-part imaging optimization model. One part is a regularization term for the forward observation process, and the other part is for constraining the sparsity of structures in the gradient domain. Using the methodology of learning to optimize, we design basic neural network modules and stack them in a cascaded manner to solve the model. We validate the proposed network using a public TomoSAR dataset. The results show that the proposed method obtains buildings with much more complete overall surfaces and more apparent edges.

The Synthetic Aperture Radar (SAR) system has undergone significant advancements, transitioning into a multi-functional platform with various modes. This study introduces a novel imaging mode that enables the simultaneous acquisition of height and intensity features, addressing the diverse requirements of different regions. Referred to as non-common band imaging, this mode optimizes bandwidth allocation within the limited illumination time, enhancing efficiency and flexibility. However, the interferometric phase retrieval problem arises in this mode. To address this challenge, we establish a forward model for the master-slave images and propose an optimization-solving model that incorporates wavelet sparsity regularization to mitigate noise interference. Furthermore, we introduce a physics-aware spectrum fusion network, combining proximal gradient descent methodology with the innovative deep unfolding technique, to restore the interferometric phase. Extensive experiments conducted on simulated and real measured data validate the effectiveness of the proposed network in terms of efficiency, accuracy, and noise reduction capabilities.

The tomographic Synthetic Aperture Radar (SAR) reconstruction based on deep learning (DL) can achieve high precision and efficiency, making it a promising method for various applications. However, current deep learning-based methods perform reconstruction by separately processing each range-azimuth resolution unit, which limits the utilization of features between resolution units and ignores the three-dimensional characteristics of the target. To address this limitation, we propose an iterative multidimensional feature reconstruction network that improves the accuracy of reconstruction by introducing correlation constraints between resolution units. Specifically, the proposed method slices the pulse-compressed data along the azimuth direction and treats each range-elevation slice as a calculation unit for reconstruction processing. And a multidimensional feature regularization module is designed to iteratively process two-dimensional features within the slice, taking into account the correlation between resolution units. Real-data experiments demonstrate that our proposed method outperforms existing network that utilize L1-norm sparse regularization in terms of achieving higher completeness.

Near-field three-dimensional synthetic aperture radar (Near-field 3D-SAR) is a powerful imaging technique with diverse applications in scattering diagnosis, person/parcel imaging, building monitoring, forest monitoring, and more. This paper provides an overview of the imaging methods employed in Near-field 3D-SAR, focusing on the underlying methodologies, imaging models, and solving flowcharts. By analyzing and summa-rizing these methodologies and flowcharts, valuable insights are uncovered. Additionally, potential research directions are identified for further exploration.

This work focuses on 3D Radar imaging inverse problems. Current methods obtain undifferentiated results that suffer task-depended information retrieval loss and thus don’t meet the task’s specific demands well. For example, biased scattering energy may be acceptable for screen imaging but not for scattering diagnosis. To address this issue, we propose a new task-oriented imaging framework. The imaging principle is task-oriented through an analysis phase to obtain task's demands. The imaging model is multi-cognition regularized to embed and fulfill demands. The imaging method is designed to be generalized, where couplings between cognitions are decoupled and solved individually with approximation and variable-splitting techniques. Tasks include scattering diagnosis, person screen imaging, and parcel screening imaging are given as examples. Experiments on data from two systems indicate that the proposed framework outperforms the current ones in task-depended information retrieval.

Video synthetic aperture radar (Video SAR) has drawn much attention because it can continuously observe and track the moving target. Rather than tracking the target directly, it is better to track its shadow because the shadow has no location shift, and the back-scattering characteristic is stable. However, most current shadow tracking methods not only suffer from false alarms because their discrimination capacities are not good enough but also suffer from missed detection because the feature extraction capacities are limited under complicated environment. Therefore, we propose a shadow-enhanced self-attention and anchor-adaptive network (SE-SA-AAN) to achieve accurate moving target tracking for Video SAR. Firstly, the pre-processing technique sparse low-rank noise decomposition (SLRND) is proposed for enhancing shadows’ salience to facilitate subsequent feature extraction. Secondly, the transformer self-attention mechanism (TSAM) is embedded in the parameters-shared backbone in the feature extraction network to concentrate on regions of interests for suppressing clutter interference. Then, the representative features are input to the detector and tracker. The detector adds the semantic guided anchor-adaptive mechanism (SGAAM) to obtain optimized anchors that match the shadows’ location and shape in each frame. Meanwhile, the tracker applies a Siamese network to achieve trajectory tracking for each shadow. Based on the detection and tracking results, a data association is applied to achieve moving targets tracking. Finally, experiments on Sandia National Laboratories (SNL) data demonstrate that SE-SA-AAN outperforms the state-of-the-art methods FairMOT, TransTrack and Centertrack by 6.4%, 7.8% and 8.3% multiple object tracking accuracy (MOTA) separately.

Tomographic synthetic aperture radar (TomoSAR) imaging algorithms based on deep learning can effectively reduce computational costs. The idea of existing researches is to reconstruct the elevation for each range-azimuth cell in one-dimensional using a deep-unfolding network. However, since these methods are commonly sensitive to signal sparsity level, it usually leads to some drawbacks like continuous surface fractures, too many outliers, et al. To address them, in this paper, a novel imaging network (AETomo-Net) based on multi-dimensional features is proposed. By adding a U-Net-like structure, AETomo-Net performs reconstruction by each azimuth-elevation slice and adds 2D features extraction and fusion capabilities to the original deep unrolling network. In this way, each azimuth-elevation slice can be reconstructed with richer features and the quality of the imaging results will be improved. Experiments show that the proposed method can effectively solve the above defects while ensuring imaging accuracy and computation speed compared with the traditional ISTA-based method and CV-LISTA.

Target areas of low resolution SAR images usually have blurred edge and large background noise, so most common object detection methods based on deep learning have obvious errors in this occasion. In this paper, we propose a high precision and lightweight network for low resolution SAR image detection. We take generalized distribution to model bounding box in training and predicting to better indicate target area boundaries in low resolution SAR images, improving detection accuracy. Moreover, we introduce “teacher-student” knowledge distilling method, which greatly reduces model parameters and further enhances the detection accuracy. Compared with conventional deep learning net-works(Faster R-CNN, SSD, CenterNet, FCOS and YOLOv3) on low resolution SAR images, the results show that our method has not only the best performance in target area extractionn, but rather light weight.

The existing Video Synthetic Aperture Radar (ViSAR) moving target shadow detection methods based on deep neural networks mostly generate numerous false alarms and missing detections, because of the foreground-background indistinguishability. To solve this problem, we propose a method to suppress complicated background of ViSAR for moving target detection. In this work, the proposed method is used to suppress background; then, we use several target detection networks to detect the moving target shadows. The experimental result shows that the proposed method can effectively suppress the interference of complicated back-ground information and improve the accuracy of moving target shadow detection in ViSAR.

Interferometric Synthetic Aperture Radar (InSAR) Imaging methods are usually based on algorithms of match-filtering type, without considering the scene's characteristic, which causes limited imaging quality. Besides, post-processing steps are inevitable, like image registration, flat-earth phase removing and phase noise filtering. To solve these problems, we propose a new InSAR imaging method. First, to enhance the imaging quality, we propose a new imaging framework base on 2D sparse regularization, where the characteristic of scene is embedded. Second, to avoid the post processing steps, we establish a new forward observation process, where the back-projection imaging method is embedded. Third, a forward and backward iterative solution method is proposed based on proximal gradient descent algorithm. Experiments on simulated and measured data reveal the effectiveness of the proposed method. Compared with the conventional method, higher quality interferogram can be obtained directly from raw echoes without post-processing. Besides, in the under-sampling situation, it's also applicable.

Inevitable interferences exist for the SAR system, adversely affecting the imaging quality. However, current analysis and suppression methods mainly focus on the far-field situation. Due to different sources and characteristics of interferences, they are not applicable in the near field. To bridge this gap, in the first time, analysis and the suppression method of interferences in near-field SAR are presented in this work. We find that echoes from both the nadir points and the antenna coupling are the main causes, which have the constant-time-delay feature. To characterize this, we further establish an analytical model. It reveals that their patterns in 1D, 2D and 3D imaging results are all comb-like, while those of targets are point-like. Utilizing these features, a suppression method in image domain is proposed based on low-rank reconstruction. Measured data are used to validate the correctness of our analysis and the effectiveness of the suppression method.

Images of near-field SAR contains spatial-variant sidelobes and clutter, subduing the image quality. Current image restoration methods are only suitable for small observation angle, due to their assumption of 2D spatial-invariant degradation operation. This limits its potential for large-scale objects imaging, like the aircraft. To ease this restriction, in this work an image restoration method based on the 2D spatial-variant deconvolution is proposed. First, the image degradation is seen as a complex convolution process with 2D spatial-variant operations. Then, to restore the image, the process of deconvolution is performed by cyclic coordinate descent algorithm. Experiments on simulation and measured data validate the effectiveness and superiority of the proposed method. Compared with current methods, higher precision estimation of the targets' amplitude and position is obtained.

Moving target shadows among video synthetic aperture radar (Video-SAR) images are always interfered by low scattering backgrounds and cluttered noises, causing poor detection-tracking accuracy. Thus, a shadow-background-noise 3D spatial decomposition (SBN-3D-SD) model is proposed to enhance shadows for higher detection-tracking accuracy. It leverages the sparse property of shadows, the low-rank property of backgrounds, and the Gaussian property of noises to perform 3D spatial three-decomposition. It separates shadows from backgrounds and noises by the alternating direction method of multipliers (ADMM). Results on the Sandia National Laboratories (SNL) data verify its effectiveness. It boosts the shadow saliency from the qualitative and quantitative evaluation. It boosts the shadow detection accuracy of Faster R-CNN, RetinaNet and YOLOv3. It also boosts the shadow tracking accuracy of TransTrack, FairMOT and ByteTrack.

In recent years, 3-D synthetic aperture radar (SAR) imaging has proved its great potential in monitoring, security inspection, and radar cross section (RCS) measurement. However, 3-D SAR images based on matched filter (MF) methods have high sidelobes and are susceptible to background noise. Therefore, in this article, we propose a novel 3-D sparse SAR imaging method to improve the image quality, which combines the plug-and-play framework and the improved alternating direction method of multiplier (ADMM). First, the plug-and-play framework allows one to use state-of-the-art denoisers instead of proximal operators to improve the image quality. Second, we linearize the subproblem of ADMM involving forward imaging model. Compared with the traditional ADMM method, the improved ADMM, namely, linear ADMM (LADMM), avoids the inversion of high-dimensional matrix and is more suitable for solving high-dimensional imaging problems. Simulation and real data experiments show that the proposed method can effectively improve the image quality. Numerical analysis and 3-D visualization results are presented, which prove the impressive performance of plug-and-play LADMM.

Ship classification in synthetic aperture radar (SAR) images is a fundamental and significant step in ocean surveillance. Recently, with the rise of deep learning (DL), modern abstract features from convolutional neural networks (CNNs) have hugely improved SAR ship classification accuracy. However, most existing CNN-based SAR ship classifiers overly rely on abstract features, but uncritically abandon traditional mature hand-crafted features, which may incur some challenges for further improving accuracy. Hence, this article proposes a novel DL network with histogram of oriented gradient (HOG) feature fusion (HOG-ShipCLSNet) for preferable SAR ship classification. In HOG-ShipCLSNet, four mechanisms are proposed to ensure superior classification accuracy, that is, 1) a multiscale classification mechanism (MS-CLS-Mechanism); 2) a global self-attention mechanism (GS-ATT-Mechanism); 3) a fully connected balance mechanism (FC-BAL-Mechanism); and 4) an HOG feature fusion mechanism (HOG-FF-Mechanism). We perform sufficient ablation studies to confirm the effectiveness of these four mechanisms. Finally, our experimental results on two open SAR ship datasets (OpenSARShip and FUSAR-Ship) jointly reveal that HOG-ShipCLSNet dramatically outperforms both modern CNN-based methods and traditional hand-crafted feature methods.

For the velocity synthetic aperture radar (VSAR) system, the velocity spectrum is obtained along the multichannel direction to separate multiple moving targets and to estimate their velocities, which is important for applications like traffic flows monitoring. On urban roads, vehicles usually appear closely with minor velocity difference, making them overlaid in the spectrum, resulting in that multiple overlaid targets are wrongly recognized as one target. To solve it, we present a multiple-overlaid-targets separation and high precision velocity estimation approach based on Bayesian criterion. Therein, velocity spectrum is reconstructed utilizing its sparsity, where the targets are separated. And their velocities are estimated precisely afterwards. Compared with the current popular approach based on MUSIC algorithm, numerical experiments show that, via the proposed one, three overlaid moving targets are separated clearly, with higher velocity resolution being obtained. And all their velocities are estimated with high precision, even in the low SNR case.