Due to its miniature size and single-pass nature, Unmanned Aerial Vehicle (UAV)-borne array Synthetic Aperture Radar (SAR) is capable of obtaining three-dimensional (3D) electromagnetic scattering information with low cost high efficiency, making it widely applicable in various fields. However, the limited payload capacity UAV platform results a number antennas affects 3D resolution. This paper proposed imaging method for UAV-borne SAR based on nested difference co-array azimuth multi...

Due to its miniature size and single-pass nature, Unmanned Aerial Vehicle (UAV)-borne array synthetic aperture radar (SAR) is capable of obtaining three-dimensional (3D) electromagnetic scattering information with a low cost high efficiency, making it widely applicable in various fields. However, the limited payload capacity UAV platform results number antennas affects 3D resolution. This paper proposes imaging method for UAV-borne SAR based on nested difference co-arrays azimuth...

Interferometric phase filtering is an indispensable step to obtain accurate measurement of digital elevation model and surface displacement. In the case low-correlation or complicated topography, traditional methods fail in balancing noise elimination preservation, which leads inaccurate interferometric phase. A new nonlocal method taking advantage higher order singular value decomposition (HOSVD) proposed this letter. For each pixel phase, a 3-D data array established, shrinkage applied...

In the complex electromagnetic environment, radio frequency interference (RFI) from other radiation sources often conflicts with synthetic aperture radar (SAR) systems, which overlaps and destroys useful data in same band, causing adverse impact to quality of SAR imaging. When faced wideband or mixed complicated RFI, traditional methods inevitablely damage original signal, cannot effectively protect reconstruct information. Besides, current semi-parametric algorithms have large computations...

Existing low-Earth-orbit synthetic aperture radar (SAR) algorithms generally assume that the data are azimuth invariant. However, this assumption does not hold for medium-Earth-orbit (MEO) SAR systems due to significantly longer integration time and complex imaging geometries. As a result, MEO cannot be processed accurately efficiently using existing algorithms. To solve problem, letter proposes two-step perturbation (AP) method uses first-step AP remove bulk variance at range processing...

Cross-modal remote sensing (RS) image retrieval aims to retrieve RS images using other modalities (e.g., text) and vice versa. The relationship between objects in the is complex, i.e., distribution of multiple types uneven, which makes matching with query text inaccurate, then restricts performance retrieval. Previous methods generally focus on feature rarely model relationships features image. Hypergraph (hyperedge connecting vertices) an extended structure a regular graph has attracted...

Compared with current low-Earth-orbit synthetic aperture radar (SAR), geosynchronous SAR (GEO SAR) is featured its ultrahigh orbit and ultralong integration time. In this letter, we answer the question whether orbital perturbation items have effects on GEO imaging during long time and, if so, how they produce errors. To achieve these goals, first develop a perturbing elements errors model based analysis. Then, propose an accurate analytical expression of to fourth Doppler parameters for...

Semantic comprehension aims to reasonably reproduce people's real intentions or thoughts, e.g., sentiment, humor, sarcasm, motivation, and offensiveness, from multiple modalities. It can be instantiated as a multimodal-oriented multitask classification issue applied scenarios, such online public opinion supervision political stance analysis. Previous methods generally employ multimodal learning alone deal with varied modalities solely exploit solve various tasks, few unify both into an...

Synthetic aperture radar automatic target recognition (SAR ATR) has been suffering from the insufficient labeled samples as annotation of SAR data is time-consuming. Thus, adding unlabeled into training attracted attention researchers. In this letter, a semi-supervised method based on consistency criterion, domain adaptation and Top-k loss proposed to alleviate need for samples. According criterion that generated by weak strong augmentations same sample belong category, we use augmented...

This paper presents a novel efficient method for gridless line spectrum estimation problem with single snapshot and sparse signals, namely the gradient descent least squares (GDLS) method. Conventional (a.k.a. measure vector or SMV) methods either rely on smoothing techniques that sacrificing range and/or azimuth resolution, adopt sparsity constraint utilize compressed sensing (CS) by defining prior grids resulting in off-grid problem. Recently emerged atomic norm minimization (ANM) achieved...

Over the past decade, rapid landscape pattern change has taken place in many arid and semi-arid regions of China, such as Yellow River Basin. In this paper, evolution was investigated by combined use satellite remote sensing, geographic information system (GIS) modelling technologies. The aim to improve our understanding changes so that sustainable land could be established. First, various metrics were analysed using structure analysis programme. Second, mathematical methodology explored...

Synthetic Aperture Radar three-dimensional imaging enables the acquisition of more comprehensive information, making it a recent hotspot in radar imaging. Traditional methods have evolved from two-dimensional and interferometric imaging, combining elevation aperture extension with signal processing techniques. limitations such as long or complex system its mechanism restrict application. In years, rapid development artificial intelligence has led to swift advancement radar, injecting new...

A novel method of interferometric synthetic aperture radar phase filtering that combines empirical mode decomposition (EMD) with Hölder exponent adjustment is presented in this letter. First, intrinsic functions (IMFs) different levels are obtained by decomposing the real and imaginary parts noisy complex formulation respectively employing EMD, which a totally data-driven without parameters to be selected. Then, we increase exponents every IMF appropriate extent according features signal...

Unsupervised methods play an essential role in Polarimetric SAR (PolSAR) image classification, where labeled data are difficult to obtain. However, there is still a large gap between existing unsupervised learning and supervised methods. Without the semantic constraints of data, pixels within same category often misclassified into different categories, leaving output be messy. To address previous issue, this paper proposes fully pipeline for training CNNs. The combines low-level superpixels...

SAR tomography (TomoSAR) has the three-dimensional (3-D) resolving ability. Most existing 3D imaging methods for TomoSAR consider layovers different channels are same. However, in low-altitude airborne cases, variation of between becomes unneglectable. In this letter, we studied channel migration phenomenon sparse under scenarios. We derived model differential range from a particular target to antennas, which is nearly linear function along array dimension. Therefore, applied Keystone...

The fast developed and widely used radar system techniques call for novel solutions on hardware design. Under a massive data source the real-time requirements in signal processing scenarios (e.g., Synthetic Aperture Radar (SAR)), reconfigurable dataflow architecture is considered as promising solution. In this article, we first extract typical algorithms systems characterize their manners. Then proposed, with high width SIMD extended PE. application implementations mapping strategy workflow...

The existence of Radio Frequency (RF) Interference will cause an adverse effect on the interpretation Synthetic Aperture Radar (SAR) images. There are various types interference, and their pattens in images vary different situations. Previous algorithms have disadvantages low precision large amount computation. In this paper, we propose a prior-induced deep neural network. Based sparse low-rank properties interference signals time-frequency domain, suppression network is designed to...

Recent advancements in generative models have provided promising solutions for synthesizing realistic driving videos, which are crucial training autonomous perception models. However, existing approaches often struggle with multi-view video generation due to the challenges of integrating 3D information while maintaining spatial-temporal consistency and effectively learning from a unified model. We propose DriveScape, an end-to-end framework multi-view, condition-guided generation, capable...

In this paper, we analyze the measurements of normalized radar cross-section(NRCS) in Wave Mode for Chinese C-band Gaofen-3(GF-3) synthetic aperture (SAR). Based on 2779 images from GF-3 quad-polarization SAR and collocated wind vectors ERA-Interim, verify feasibility using ocean surface fields VV-polarized NRCS to perform calibration. The method uses well-validated empirical geophysical model function (CMOD4) estimate calibration constant each beam. Amazon rainforest experiment results show...

On 23 November 2021, the Gaofen-3-02 (GF3-02) satellite was successfully launched in Jiuquan Satellite Launch Center of China. The primary payload is C-band Synthetic Aperture Radar (SAR), with a maximum resolution 1 m, and includes 12 imaging modes such as Spotlight, Strip, TOPSAR, which will play an essential role marine environment monitoring. As important environmental parameter, wind speed accuracy retrieved by GF3-02 SAR directly reflects its performance effectiveness operational...