Plug-and-Play Synthetic Aperture Radar Image Formation Using Deep Priors

Alver, Muhammed Burak; Saleem, Ammar; Çetin, Müjdat

doi:10.1109/tci.2020.3047473

Cited by 39 publications

(20 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The convolutional neural network (CNN) utilized in Hybrid-WASAR is pre-trained following a similar approach as described in ref. [18]. We utilized the Gotcha Public Data Products for conducting the experiment.…”

Section: Experiments and Analysis Of Resultsmentioning

confidence: 99%

Wide angle SAR imaging method based on hybrid representation

Zhao

Chen

矢田

et al. 2023

Electronics Letters

View full text Add to dashboard Cite

In this paper, the application of a hybrid representation in wide‐angle synthetic aperture radar (WASAR) imaging is investigated, addressing the challenge of achieving sparse representation in the presence of complex electromagnetic scattering characteristics and highly anisotropic targets. A convolutional neural network is utilized to represent the two‐dimensional data within each azimuth aperture, while employing blind compressed sensing (BCS) to achieve sparse representation across different azimuth apertures. Convolutional neural network (CNN) excels in learning spatial hierarchies and local dependencies of two‐dimensional data but requires a large amount of training data. Isotropic targets within each azimuth aperture can be used for training conventional SAR data, while acquiring training samples for anisotropic targets poses challenges. To address this issue, BCS are integrated into WASAR imaging, eliminating the need for additional training data. By integrating these methods, a novel approach called hybrid‐WASAR is proposed, which incorporates two regularization terms into WASAR imaging and employs the alternating direction method of multipliers to iteratively solve the imaging model. Compared to traditional WASAR imaging techniques, hybrid‐WASAR improves the accuracy of reconstructing the backscattering coefficients of the targets, leading to a significant enhancement in overall imaging quality.

show abstract

Section: Experiments and Analysis Of Resultsmentioning

confidence: 99%

Wide angle SAR imaging method based on hybrid representation

Zhao

Chen

矢田

et al. 2023

Electronics Letters

View full text Add to dashboard Cite

show abstract

“…Inspired by this, we find it is enough to construct a simulated dataset to mimic the deterioration manner of down-sampling. Similar strategies have been adopted in some pioneer works like [40] as well. The simulated echoes are generated under the premise of knowing imaging geometry.…”

Section: Training Details and Imaging Implementation 1)mentioning

confidence: 96%

“…4 illustrates the whole pipeline of network training and implementation of 3-D holography via LFIST-Net, where S raw and S 3D denote the collected sparsely sampled echo cube and range focused cube, respectively. Mathematically, the reconstruction process for range cell n can be expressed as S 3D = FT range (S raw ) α 3D (n, :, :) = F † (LFIST(S 3D (n, :, :))) parallelly process in range direction (40) where FT range (•) represents Fourier transformation in the range direction. LFIST(•) denotes the well-trained LFIST-Net model.…”

Section: Training Details and Imaging Implementation 1)mentioning

confidence: 99%

See 1 more Smart Citation

Lightweight FISTA-Inspired Sparse Reconstruction Network for mmW 3-D Holography

Wang

Wei

Liang

et al. 2022

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Integrating compressed sensing (CS) with millimeter-wave (mmW) holography has shown great potential to achieve lightweight onboard hardware, low sampling ratio, and high-speed sensing. However, conventional CS-driven algorithms are always limited by nontrivial adjusting of parameters and excessive computational cost caused by plenty of iterations. To address this problem, we propose a lightweight model-based deep learning framework (LFIST-Net) for mmW 3-D holography, by combining the interpretability of fast iterative shrinkage-thresholding algorithm (FISTA) and tuning-free merit of data-driven deep neural network. First, the single-frequency (SF) holographic imaging technique is integrated into FISTA, which serves as the sensing kernels, to avoid large-scale matrix multiplications. Subsequently, the kernel-based FISTA (KFISTA) is mapped into layer-fixed and parameter-learnable LFIST-Net, whose weights are relaxed to be layer-varied. The updating of key parameters in LFIST-Net, including step sizes, thresholds, and momentum coefficients, are regularized by soft-plus function to ensure the non-negativity and monotonicity. As for 3-D holography implementation, the "1-D + 2-D" scheme is adopted, where the matched filtering (MF) and well-trained LFIST-Net are used for range focusing and reconstructions of azimuth slices. Without losing efficiency, the range-focused subechoes are processed parallelly in 3-D cube form. Experiments, including both simulated and measured tests based on a commercial mmW radar, prove that LFIST-Net is capable of reconstructing the imaging scene precisely. In particular, in near-field mmW 3-D holography tests, both numerical and visual results demonstrate LFIST-Net yields compelling reconstruction performance while maintaining high computational speed compared with MF-based, conventional CS-driven, and network-based methods. Index Terms-3-D holography, compressed sensing (CS), deep unfolding, fast iterative shrinkage-thresholding algorithm (FISTA), millimeter-wave (mmW) imaging, near-field, sparse reconstruction. I. INTRODUCTIONN EAR-FIELD 3-D millimeter-wave (mmW) holography has gained significant concerns in the microwave theory and techniques society. mmW, possessing the merits of reasonable penetration for most nonmetal materials and no known Manuscript

show abstract

“…Synthetic aperture radar (SAR) is a technique that is receiving more and more attention worldwide nowadays, given that it is capable of imaging the observed scenes of interest, regardless of weather conditions, both day and night [1][2][3][4][5][6]. SAR has been widely employed in different applications, such as environmental observation, traffic monitoring, sea-surface surveillance, and antiterrorism.…”

Section: Introductionmentioning

confidence: 99%

Fast Approach for SAR Imaging of Ground-Based Moving Targets Based on Range Azimuth Joint Processing

et al. 2022

View full text Add to dashboard Cite

The synthetic aperture radar (SAR) images of a moving target may be out of focus, given the motions of a non-cooperative target. Doppler ambiguities, including the Doppler center blur and spectrum ambiguity, will easily appear due to the limitations of pulse repetition frequency, which causes difficulty in moving-target imaging. Therefore, a robust fast Doppler ambiguity approach for SAR imaging of a ground-based moving target using range azimuth joint processing (RAJP) is presented. Firstly, the use of RAJP, based on a two-dimensional cross-correlation function and linear range cell migration (LRCM) compensation function, is proposed to simultaneously obtain the first- and second-order phase parameters in the fast-time and azimuth-frequency domains. Then, a corresponding azimuth reference function is constructed to image the moving target. Additionally, a principal component analysis-based operation is introduced to solve the mismatch with the LRCM compensation function. The couplings between the range and azimuth and between the first- and second-order parameters can be simultaneously decoupled by the proposed RAJP operation, which simplifies the processing steps. The developed approach can simultaneously obtain the first- and second-order parameters in the fast-time and azimuth-frequency domains, which avoids the propagation error of parameter estimation caused by the stepwise processing operation. The proposed method is relatively fast, given the need for fewer processing steps. The presented approach is robust in terms of Doppler ambiguity and handles the blind speed sidelobe well. In this study, simulated and real data are processed to verify the proposed approach.

show abstract

Plug-and-Play Synthetic Aperture Radar Image Formation Using Deep Priors

Cited by 39 publications

References 52 publications

Wide angle SAR imaging method based on hybrid representation

Wide angle SAR imaging method based on hybrid representation

Lightweight FISTA-Inspired Sparse Reconstruction Network for mmW 3-D Holography

Fast Approach for SAR Imaging of Ground-Based Moving Targets Based on Range Azimuth Joint Processing

Contact Info

Product

Resources

About