Synthetic aperture radar (SAR) tomography (TomoSAR) technology effectively provides precise three-dimensional (3D) forest vertical structure information. However, conventional TomoSAR methods require abundant acquisitions for accurate 3D reconstruction, which is time-consuming and low-efficiency for forest vertical structure mapping. To address these limitations, this paper proposes a novel micro-stack TomoSAR imaging approach utilizing four images, referred to as the double iterative adaptive residual approach (DIARA). The DIARA innovatively combines iterative inner-outer adaptive spectral estimation and residual optimization to enhance both processing efficiency and accuracy. For validation purposes, both simulated and airborne TomoSAR experiments are analyzed at P- and L-band. The P-band results from the BorTomoSAR campaign indicate that the DIARA acquires higher accuracy for estimating forest height than the conventional methods, i.e., improvingR2from 0.443 to 0.628, mean absolute error (MAE) from 0.779 m to 0.625 m, mean absolute percentage error (MAPE) from 4.629% to 3.680%, and root mean square error (RMSE) from 0.941 m to 0.771 m. Additionally, the performance is further validated by the TropiSAR P-band campaign, which confirms the robustness of the proposed DIARA in high-canopy, complex forest environments. Furthermore, the L-band results from HaiTomoSAR campaign demonstrate that the DIARA method successfully detects the weak ground scatterers beneath dense forest canopies, which validates its super-resolution capability in vertical structure reconstruction.

Airborne Synthetic Aperture Radar (SAR) Tomography (TomoSAR) technology facilitates the extraction of three-dimensional (3D) information of target scatterers. However, phase screens induced by radar platform trajectory errors causes TomoSAR defocusing, which adversely affects the accuracy of the forest vertical structure retrieval. Additionally, the phase screens exhibit space-variant characteristics, which significantly degrade the calibration performance estimated by traditional phase gradient autofocus (PGA). This paper proposes an improved phase calibration method synthesizing unconstrained optimization model and weighted PGA (WPGA), which effectively address the above challenges. Firstly, the SAR data stack is segmented into multiple subareas by assuming that phase screens are space-invariant within each small area, which reduces complexity and improves computational efficiency. Secondly, a WPGA method synthesizing the scatterer heights derived from an unconstrained optimization model is proposed to estimate the phase screens. Simulation experiments are conducted to validate the effectiveness of the proposed phase calibration method. Furthermore, the full-polarization SAR data stacks acquired by the BorTomoSAR campaign are used for tomographic focusing analysis. Experimental results demonstrate that the proposed method accurately estimates the phase screens at the P, L, and S bands, providing a efficient solution for the forest vertical structure retrieval.

Segmentation is an essential part in SAR image analysis and application. However, the presence of speckle noise and intensity inhomogeneity regions leads to great difficulty in SAR image segmentation. In this letter, we propose a novel level set method that integrates the local intensity and the global feature information for better segmentation results. To establish the energy function, a new image model is developed to describe the inhomogeneous SAR images and form the local intensity clustering term. Meanwhile, the feature map extracted by invariant moment is used to define the global fitting term through a piecewise constant function. Segmentation is realized by solving the level set function from the minimization of the energy function. Experiments on the synthetic and real SAR images verify the effectiveness of the proposed method.

This letter proposes a new semantic segmentation network called IGFUnet to map the urban areas from large scale SAR images. The network is based on U-net structure which is especially suitable for semantic segmentation of small data set. Inception module is adopted as the encoder to extract rich and dense features. We develop a global feature attention module (GFA) as a channel attention module to perform feature correction on each decoding layer. Combing the inception module and GFA module can take full use of the feature map and efficiently integrate features from different scales. Qualitative and quantitative experiments based on Sentinel-1A have proved the effectiveness of our IGFUnet model.

Clutter suppression is a key step for efficient detection of moving targets and accurate estimation of their parameters. Current clutter suppression approaches are available, in the case that clutter echoes of each channel are free of Doppler ambiguity. However, for multichannel high-resolution wide-swath (HRWS) synthetic aperture radar (SAR) system, the received echoes of each channel suffer Doppler ambiguity and thus current clutter suppression approaches may not perform well. To address this issue, the signal models of stationary and moving target with Doppler ambiguity are derived. By analyzing the signal models, this paper proposes a clutter suppression approach for multichannel SAR system with the capability of HRWS imaging. Simulated results and real data results both demonstrate the validity of the proposed approach.

This letter focuses on a modified complex-valued Markov random field (CMRF) modeling filter to improve the performance in the filtering of the synthetic aperture radar interferometric phase. In the reference CMRF modeling phase map filter, the CMRF model was employed to update residues and their neighbors. By this, the residues were reduced, and phase jumps were preserved simultaneously. However, residue reduction was still insufficient. Furthermore, incorrect model parameter estimation leads to wrong filtering in some blocks. To solve the two aforementioned problems, we propose a modified CMRF modeling filter, where the original interferogram is divided into overlapped blocks. In addition, the adaptive weighted neighbor values are used to estimate the CMRF model parameters. Both simulated and real data experiments are performed to validate this method.

In this letter, an approach based on spectral diversity and conventional Interferometric Synthetic Aperture Radar (InSAR) for Digital Elevation Model (DEM) reconstruction is proposed. Conventional InSAR technique often fails to correctly reconstruct the DEM due to the presence of discontinuities and/or interferometric noise. With the assistance of the spectral diversity method, the ill-posed phase-unwrapping can be regularized. Thus the capabilities of the conventional InSAR technique for DEM reconstruction can be notably improved. Comparing to the multi-channel InSAR technique, the proposed method requires less SAR images. Comparing to the multifrequency technique, the performance of the proposed method depends on the perpendicular or effective baseline and the ratio of the radar's carry frequency to the bandwidth between the central frequencies of the two sub-looks; while the multifrequency technique depends more on the range bandwidth. Thus, the proposed method is more flexible than the multifrequency technique. The effectiveness of the proposed algorithm is validated by experimental results using real data.

In this paper, an adaptive patch-based filter for Synthetic aperture Radar (SAR) Interferometric phase in the Fourier transform complex domain is proposed. The proposed filter fully utilizes the properties of locally planar terrain phase noise model so that the noise can be suppressed according to the local level of noise. Moreover, the denoising performance of the proposed filter is further improved by an adaptive selection strategy for patch size. The experimental results show that the proposed method delivers state-of-the-art denoising performance in terms of objective criteria.

Conventional frequency-domain frequency-modulation continuous-wave (FMCW) synthetic aperture radar (SAR) raw data simulators (RDSs) can only handle the case of static targets while the scene with moving targets cannot be accommodated. In FMCW SAR, the motion of the radar and moving targets during the pulse duration must be considered, thus posing a challenge for the FMCW SAR RDS. An accurate point target reference spectrum that considers this motion is first derived in the literature. Based on the spectrum, we design an accurate and efficient FMCW SAR RDS. Compared with its counterparts, it can precisely simulate the raw data of moving targets even in the case of high resolution and high squint. Both point target and extended scene simulations are performed to validate the simulator, and its efficiency is also analyzed. Results show its outstanding performance for generating the FMCW SAR raw data of extended scenes with static and moving targets.

This letter presents an accurate range model, named as Doppler parameter range model (DRM), to perform signal processing for geosynchronous synthetic aperture radar (Geo-SAR). Besides that a credible approach, defined as accurate coordinate numeric model (ACNM), is proposed to simulate the raw echo signals. To improve the accuracy of DRM and ACNM, the rotating Earth is modeled as an ellipsoid and the eccentricity of the satellite's orbit is analyzed. The superiority of DRM that it is more precise than hyperbolic range equation (HRE) and advanced hyperbolic range equation (AHRE), presented before, is illuminated. The point target 2-D frequency spectrum of DRM is derived. Finally, the simulation results validate that DRM and ACNM has a good effect on an L-Band Geo-SAR system with an azimuth resolution around 5 m.

As a novel mode of ground-based synthetic aperture radar (GB-SAR), Arc frequency-modulated continuous wave (FMCW) SAR is rarely discussed in the literature up to now. Compared with the conventional rail GB-SAR system, the synthetic aperture is formed by the rotation of antennas, which can scan a wide azimuth extent. Due to its convenience and potential in SAR applications, the Institute of Electronics, Chinese Academy Sciences, carried out a series of Arc FMCW SAR experiments in March 2013, and a lot of results were obtained. In this paper, we discuss the signal processing of Arc FMCW SAR and its first results in change detection and interferometric SAR. The components of the Arc FMCW SAR system are first described, and then, we focus on its signal processing, where the signal model and its properties are investigated; two focusing algorithms are developed for different purposes. The effectiveness of the algorithms is validated by both simulation and real data experiments. We also exhibit the applications of Arc FMCW SAR in landslide detection and digital-elevation-model extraction for the first time. Results show its huge potential in ground-based applications.

High-resolution spaceborne synthetic aperture radar (SAR) mainly poses two challenges to signal processing. The first challenge involves the signal model, where a precise range equation of spaceborne SAR should be considered as the conventional hyperbolic range equation fails to precisely describe the range history in the high-resolution case. The second challenge is an efficient focusing algorithm since the existing SAR processors are inaccurate or inefficient for high-resolution spaceborne SAR. Therefore, in this paper, a novel fourth-order polynomial range equation based on Doppler parameters is proposed, and the method for parameter determination is also addressed. Compared with conventional range equations, the presented one is more accurate and concise for low-earth-orbit SAR so that a higher azimuth resolution can be achieved. Based on the range model, a 2-D spectrum is derived, and an extended range Doppler domain algorithm for SAR image formation in the sliding spotlight mode is also developed. Additionally, we carried out several simulations to validate the presented approach. Results demonstrate high performances of the focusing algorithm as well as the range equation.

Besides the evident superiority, the Geosynchronous Earth orbit synthetic aperture radar (Geo-SAR) system meets great more challenges than the current Low-Earth-Orbital SAR (Leo-SAR) and the Medium-Earth-Orbit SAR (Meo-SAR). The signal processing approach for spotlight mode Geo-SAR is discussed in this paper. A new range model called the sixth-order Doppler parameter range model (DRM-6) is derived based on the accurate sixth-order Doppler parameter calculation algorithm. The accuracy of Hyperbolic Range equation (HRE) and the DRM-6 are compared. According to the non-linear property of the synthetic aperture, the Back projection algorithm is chosen for image formation. At last the simulation results validate that DRM-6 has a good effect on the L-Band spotlight Geo-SAR system with a azimuth resolution around 2m.

In this paper, we present a modified frequency scaling processing method for FMCW SAR data. Compared to the conventional frequency scaling algorithm for FMCW SAR, an analytical slant model is employed so that it can work well in all conditions for FMCW SAR. Firstly, the algorithm model is deduced in detail and its implementation is given, and then a simulation experiment is carried on to evaluate its efficiency. Finally, real data of FMCW SAR validate the proposed algorithm.

The range equation employed in the current low earth orbital Synthetic Aperture Radar (Leo-SAR) systems is the hyperbolic range equation (HRE). The key parameters of it are effective velocity and effective squint angle. Traditionally the calculation method of the two values may lead to large errors, and it is only fit for Leo-SAR with moderate space resolution. To get the azimuth resolution on the order of one meter or even higher, an improved method based on the accurate second-order Doppler parameters is used in this paper to decrease the errors of the two parameters above. Then we can get the improved hyperbolic range equation (IHRE). The theory is validated by the parameters of TerraSAR-X, and the image with the space resolution around 0.35m is focused perfectly.

This paper presents a novel agent architecture, the VM-architecture, that chooses its actions via a behaviour instantiation mechanism inspired by Maes?s behaviour networks. The VM-architecture enables an agent, VMattie, built upon it, actively to gather information from humans using email in natural language, to compose announcements of next week?s seminars and to mail them each week to a list that she keeps updated. VMattie does all this without the supervision of a human. Implemented and tested in a UNIX environment, the system?s performance surpassed the authors? fondest hopes.