No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Understanding Synthetic Aperture Radar Images
... However, their use is limited by cloud cover and illumination conditions. SAR does not suffer from these limitations, yet SAR data from each individual space mission are tipically single-frequency with 1to-4 channels (depending on polarimetry) [2], a scenario that limits their capability to discriminate land-covers. From this perspective, fusing data obtained from distinct SAR missions is particularly promising for land cover mapping, since it combines all-weather and day-and-night acquisition with the M. Pastorino, G. Moser, and S. B. Serpico are with the DITEN Department, University of Genoa, 16145, Genoa, Italy (e-mail: martina.pastorino@edu.unige.it). ...
... the support is gratefully acknowledged. possibility to jointly use multiple observations associated with distinct carrier frequencies and spatial resolutions [2]. In this framework, the challenge is to jointly exploit the heterogeneous multimodal information conveyed by multimission SAR to produce accurate semantic segmentation maps [3]. ...
... On the other hand, the ensemble and deep learning techniques integrated in the developed methodology, thanks to their non-parametric formulation, make it possible to incorporate data with arbitrary probability distributions, thus effectively supporting the application to SAR data associated with both textured image areas, typically following K [2], Fisher [11], generalized Gaussian Rayleigh [12], and generalized Gamma distributions [13], and with areas without texture, usually well modeled by Gamma, Weibull or log-normal distributions [2]. ...
The availability of multimodal remotely sensed images calls for the development of methods capable to jointly exploit the information deriving from images acquired at different spatial resolutions, frequencies, and bands, taking advantage from their possible complementary features. This paper proposes to address this task in the case of multimission synthetic aperture radar (SAR) images, through a combination of fully convolutional networks (FCNs), hierarchical probabilistic graphical models (PGMs), and decision tree ensembles. The objective is to model the multimodal information collected at multiple spatial resolutions by distinct space missions with SAR payloads through the non-parametric formulation of FCNs and decision trees, and the spatial and multiresolution modeling capabilities of FCNs and hierarchical PGMs. The experimental validation is conducted with multimission SAR imagery acquired at X-, L-, and C-band respectively by COSMO-SkyMed, SAOCOM, and Sentinel-1 over Northern Italy. The results suggest the advantages of incorporating multifrequency radar acquisitions to reach accurate classification maps, and the multimodal fusion capabilities of the proposed methodology.
... Furthermore, clutter data with a certain degree of heterogeneity can be addressed by adjusting the value of κ. However, according to Oliver and Quegan [26], the Weibull model fails at describing multilook clutter. ...
... The Rayleigh distribution is the basic scattering model that results from the central limit theorem (CLT) [26]. Invoking the CLT, the complex backscattered signal is a bivariate Normal random variable with i.i.d. ...
... This distribution is only adequate to model single-look clutter amplitude data from homogeneous areas [26]. However, the amplitude of high resolution (HR) SAR clutter data often exhibits non-Rayleigh behavior, thus requiring a more flexible model [15]. ...
Clutter modeling is used in diverse Synthetic Aperture Radar (SAR) image processing fields, e.g., speckle suppression, detection, classification, and recognition. Therefore, accurate formulation of SAR clutter models is crucial to achieve effective performance in such applications. We present a survey on parametric and semi-parametric distributions in characterizing SAR ground clutter statistics, and we compare estimators for their parameters (when explicitly available) with a theoretical assessment of their computational burden. Furthermore, we discuss how to assess these models with the
k
3 ∼
k
2 diagram. We also discuss how to analyze the homogeneity of SAR clutter data using clutter models’ coefficient of variation (
C<sub>v</sub>
) to justify their effectiveness in portraying scene heterogeneity. Experiments are made on simulated data and SAR images to assess the goodness-of-fit, deviation of estimated
C<sub>v</sub>
from the observed
C<sub>v</sub>
, and computational complexity for the state-of-the-art clutter models, thereby assessing their effectiveness in characterizing SAR clutter amplitude statistics. The MATLAB code that implements these tools is available at https://github.com/dkmahapatra1/SAR-Clutter-Modelling.git
... Xie et al. put forward the ideas about various SAR data formats and the common distributions such as amplitude (Rayleigh), intensity (Exponential), phase (Uniform) and also explained about the different noise distributions and log transformed distributions (Xie et al., 2002). Oliver et al. detailed about the SAR technology, data formats and the radar cross section concepts as well (Oliver & Quegan, 2004). Zaart and Ziou put forward an idea about the GGBL system, which is a detailed comparison of the Gaussian, Gamma, b and Lognormal distributions. ...
... The proposed method adopts the advantages of optimized linear filters in transform domain. The SAR data taken in amplitude, intensity, phase and log intensity formats as shown in Figure 2 (Oliver & Quegan, 2004;Bhattacharya, n.d.). ...
... Sample SAR image in different data formats-(a) Real in-phase, (b) Quadrature imaginary, (c) phase, (d) Amplitude, (e) Intensity, (f) Log intensity(Oliver & Quegan, 2004). ...
Synthetic Aperture Radar (SAR) imagery finds widespread applications across engineering disciplines, encompassing vegetation surveys, biomass estimation, and weather-related investigations. However, two significant challenges hinder SAR image analysis: the inherent difficulty in visually interpreting SAR images and the adverse impact of multiplicative speckle noise. The proposed study introduces a robust framework for removing G ⁰ distributed noise from Synthetic Aperture Radar (SAR) images. This framework utilizes Bitonic preprocessing filters and SailFish optimization within the wavelet domain to address challenges in SAR image analysis, particularly the difficulty in visually interpreting SAR images and the impact of multiplicative speckle noise. By assessing noise distribution and applying denoising algorithms tailored to estimated noise distribution, the study overcomes limitations of linear filters by adopting Bitonic filters for effective preprocessing. Through parameter optimization using SailFish Optimization (SFO) and objective functions like Structural Similarity Index Measure (SSIM) and Edge Preservation Index (EPI), the proposed method outperforms conventional algorithms on both synthetic and real SAR data. Furthermore, for synthetic SAR data with noise modeled as G⁰ distributed, our approach provides a more realistic benchmark for comparison than existing methods.
... Las imágenes de radar de apertura sintética (SAR) son producidas por sensores activos, ya que el mismo sensor emite y recibe la señal electromagnética con la que obtiene información de la superficie terrestre (Trebits, 1987;Zhu et al., 2017a). En el caso de los sensores SAR, usualmente emiten una señal en la longitud de onda de las microondas (aprox. 1 -130 cm), la cual es reflejada por la superficie terrestre y medida por el sensor (Moreira et al., 2013;Oliver and Quegan, 2004). Generalmente, la información contenida en las imágenes SAR responde a características geométricas (i.e., volumen de un bosque o distribución espacial de los troncos) y de humedad de la superficie. ...
... La retrodispersión de una superficie va a estar determinada por (Sinha et al., 2015): 1) la longitud de onda de la señal, 2) la humedad de la superficie y 3) las características geométricas de la superficie. En el primer caso, la longitud de la banda del SAR (λ) va a determinar qué superficies se verán como rugosas y la capacidad de penetración de la señal en cubiertas forestales (Flores-Anderson et al., 2019;Oliver and Quegan, 2004). En general, las superficies cuya rugosidad (h) tenga una relación h < λ / 32 se verán como planas (baja retrodispersión), mientras que las que guarden una relación h > λ / 2, como rugosas (alta retrodispersión). ...
... En el segundo caso, mientras más húmeda esté una superficie, mayor será su retrodispersión (Flores-Anderson et al., 2019). Por último, de acuerdo a la polarización de la señal, se pueden distinguir tres tipos generales de respuesta (Fitch, 1988;Oliver and Quegan, 2004). El primero corresponde a superficies rugosas, donde la señal co-polarizada VV suele tener los valores más altos de retrodispersión. ...
La deforestación y la degradación forestal son dos procesos que contribuyen al cambio global mediante la emisión de gases de efecto invernadero, la pérdida de biodiversidad y la reducción en la calidad de algunos servicios ecosistémicos. Gracias a su capacidad de estudiar grandes superficies y contar con un registro histórico, la percepción remota ha demostrado ser una herramienta esencial para cuantificar estos procesos. Además, el desarrollo de nuevos métodos de análisis como los algoritmos de aprendizaje profundo e imágenes de mayor resolución espectral, espacial y temporal gratuitas abren la posibilidad de desarrollar métodos que permitan mejorar las capacidades actuales de monitoreo. En este contexto, el presente trabajo evaluó el desempeño de los algoritmos de aprendizaje profundo con imágenes multiespectrales y de radar de apertura sintética para clasificar
distintos tipos de coberturas del suelo, detectar la deforestación y la degradación forestal. Los resultados muestran que los algoritmos de aprendizaje profundo, en general, permiten obtener mejores resultados que su contraparte de aprendizaje automatizado, debido a la incorporación de patrones en las dimensiones espaciales y temporales. Sin embargo, dicho desempeño está condicionado al tamaño de muestra y la fuerza de la relación entre la señal remota y la clase o característica a evaluar. Por otro lado, la combinación de la información multiespectral y radar de apertura sintética en general fue beneficiosa, aunque en algunos casos el agregar la información radar brinda resultados similares que solo con la multiespectral. Este último resultado se debe a la falta de contribución de nueva información útil por parte de las imágenes de radar, en comparación con la que ya brindan las imágenes multiespectrales. Los resultados apuntan a que este tipo de técnicas permiten
obtener resultados más precisos para identificar estos procesos, por lo cual, representan una herramienta atractiva para el desarrollo de futuras herramientas encaminadas al monitoreo de estos procesos.
... Fully-developed speckle is characterized in the spatial frequency domain as a white process. In the complex spatial domain (in-phase and quadrature components), it is modeled as a spatially uncorrelated multiplicative random process, independent of the underlying radar reflectivity and characterized by a zero-mean circular Gaussian probability density function (PDF) [1]. During the processing of raw data to create an image, however, there may be stages that violate the validity of this assumption. ...
... A. OVERVIEW OF SAR SYSTEM SAR images are computed images obtained by processing radar echoes taken from a moving platform [1], [4]. A train of pulses, or chirps, is sent along a leaning direction across the track of the platform with respect to the vertical, referred to as slant-range. ...
... Let S w pq (r) denote the discrete complex scattering coefficient, where the subscripts pq denote the transmitting and receiving polarizations, namely, HH, VV, HV and VH, and r ≜ (r x , r y ) denotes the two-dimensional coordinates in the image plane. Under the assumption of fully-developed speckle, S w pq (r) is a zeromean, white complex circular symmetric Gaussian process, with variance σ pq (r), which is the radar reflectivity imaged by the system [1]. The complex image at the output of a SAR processor equipped with a 2D separable tapering window H (f ), with f ≜ (f x , f y ), can be formulated as a convolution of the scattering coefficient by the inverse Fourier transform of the frequency tapering window, H (f ), or point spread function (PSF), h(r): ...
This study employs an unsupervised procedure to spatially decorrelate fully-developed speckle in single-look complex (SLC) synthetic aperture radar (SAR) images. The goal is evaluating the extent to which the spatial correlation of the noise induced by the SAR processor affects the detection accuracy of temporal variations of land-cover between two one-look images of the same landscape acquired on different dates. To simulate the scenario, we have spatially correlated a synthetic map of white complex circular symmetric Gaussian noise by using a two-dimensional separable Hamming window in the Fourier domain. The correlated complex speckle field has been modulated by a noise-free optical view, to simulate an SLC SAR image. Subsequently, we have reduced the correlation of the SLC image through a whitening process and calculated the modulus of the complex image. We have applied various methods of statistical change detection for real-valued SAR data, and compared the accuracy of change maps in the following cases: i) ideally uncorrelated noise, ii) correlated noise, iii) correlated noise that has been decorrelated. The study considers three change detection algorithms, ranging from the basic Log-Ratio operator preceded by despeckling to advanced parametric and nonparametric methods based on Kullback-Leibler distance and mean-shift clustering of bivariate scatterplots of local means. Simulation results demonstrate consistent performance improvements, in terms of both geometric accuracy and reduced number of false alarms.
... Icebergs can be automatically detected by using either segmentation (Kim et al., 2011;Tao et al., 2016a;Akbari and Brekke, 2018;Karvonen et al., 2022) or global thresholding approaches (Dierking and Wesche, 2014;Barbat et al., 2019). The most common approach is the application of adaptive thresholding techniques such as the constant false alarm rate (CFAR) detector (Oliver and Quegan, 2004). CFAR detectors are especially valuable for wideswath SAR images, where large variations in incidence angles make global thresholding techniques difficult to design. ...
... The K distribution is a PDF that has been widely used to model sea surface clutter (Oliver and Quegan, 2004), and CFAR algorithms based on the K distribution have been used for ship and iceberg detection (Power et al., 2001;Brekke and Anfinsen, 2011;Wesche and Dierking, 2012;Liu, 2018). But due to the complexity of the K distribution, models based on simpler PDFs are also commonly used, e.g., the log-normal distribution (Crisp, 2004;El-Darymli et al., 2013) and gamma distribution (Gill, 2001;Crisp, 2004;Buus-Hinkler et al., 2014;Tao et al., 2016b). ...
... The gamma detector is based on the fact that, under fully developed speckle, the multi-looked background clutter intensity follows a gamma distribution (Oliver and Quegan, 2004;Argenti et al., 2013). Here, the threshold for determining outliers can then be found from the average clutter intensity and the number of looks, L, which is known. ...
In this study, we pursue two objectives: first, we compare six different “constant false alarm rate” (CFAR) algorithms for iceberg detection in SAR images, and second, we investigate the effect of radar frequency by comparing the detection performance at C- and L-band. The SAR images were acquired over the Labrador Sea under melting conditions. In an overlapping optical Sentinel-2 image, 492 icebergs were identified in the area. They were used for an assessment of the algorithms' capabilities to accurately detect them in the SAR images and for the determination of the number of false alarms and missed detections. By testing the detectors at varying probability of false alarm (PFA) levels, the optimum PFA for each detector was found. Additionally, we considered the effect of iceberg sizes in relation to image resolution. The results showed that the overall highest accuracy was achieved by applying a log-normal CFAR detector to the L-band image (F score of 70.4 %), however, only for a narrow range of PFA values. Three of the tested detectors provided high F scores above 60 % over a wider range of PFA values both at L- and C-band. Low F scores were mainly caused by missed detections of small-sized (<60 m) and medium-sized (60–120 m) icebergs, with approximately 20 %–40 % of the medium icebergs and 85 %–90 % of small icebergs being missed by all detectors. The iDPolRAD detector, which is sensitive to volume scattering, is less suitable under melting conditions.
... According to [17], if speckle is fully developed in a SAR image, statistical models of the speckle can be obtained under the assumption that the real and imaginary parts of a complex SAR echo signal are independently Gaussian-distributed with a zero mean and identical variance. In this case, negative exponential [16], Rayleigh [16], Gamma [18], and Nakagami [19] distributions have been proposed to characterize single-look intensity speckle, single-look amplitude speckle, multilook intensity speckle, and multilook amplitude speckle, respectively [20]. However, for SAR images of complex scenes, e.g., high-resolution SAR images and heterogeneous images, the above models are no longer applicable. ...
... According to [19], the set of samples obtained within the full synthetic aperture can be split into several adjacent subsets. Each subset forms a separate image, namely, a 1-look image. ...
... ). Obviously, the shadow amplitude multilooked in the real domain follows a Nakagami distribution. Similarly, the comparison with [16] indicates that Equation (19) and the PDF proposed in [16] for multilook intensity speckle are in the same form. The significant difference lies in the parameters α and σ. ...
Shadows are a special distortion in synthetic aperture radar (SAR) imaging. They often hamper proper image understanding and target recognition but also offer useful information, and therefore, the statistical modeling of SAR image shadows is imperative. In this endeavor, we systematically deduced the statistical models of shadows in multimodal SAR images, including single-look intensity and amplitude images and multilook intensity and amplitude images in a real domain and complex domain, respectively. In particular, for the filtered SAR image shadow, we introduced the generalized extreme value (GEV) distribution to characterize its statistics. We carried out an experiment on shadows in a real SAR image and conducted chi-square goodness-of-fit tests on the deduced models. Furthermore, we compared the deduced statistical models of shadows with state-of-the-art statistical models of SAR imagery. Finally, suggestions are given for selecting the optimal statistical model of shadows in multimodal SAR images.
... When a surface is illuminated by SAR radiation, scattering within a SAR resolution cell (pixel) causes energy to be reflected in the sensor with random phase shifts. Each pixel's distinct responses are combined by the receiver Oliver and Quegan (2004). The amplitude and phase of the radar returns fluctuate randomly from pixel to pixel because of the non-uniform distribution of scattered contributions. ...
... Typically, the width of this function is used to quantify the resolution. According to the fundamental principle applicable to all radar systems, the available resolution is inversely proportional to the system bandwidth Oliver and Quegan (2004); Woodhouse (2005). For side-looking radars, the resolution is defined independently in two perpendicular dimensions: the azimuth (along-track) direction and the range (across-track) direction. ...
To comprehensively comprehend the potential risks posed by seismic activity and the formation of geological structures in areas with high tectonic activity, it is critical to conduct thorough investigations on the deformation of the Earth's crust caused by earthquakes. By analyzing the changes in the crust's shape and movements resulting from seismic events, researchers can gain valuable information on the underlying geological processes and the potential for future earthquakes. This knowledge can help in developing effective strategies to mitigate the impact of earthquakes on society and infrastructure, as well as contribute to a deeper understanding of the Earth's dynamic systems. In this study, we use Sentinel-1 interferometric synthetic aperture radar (InSAR) images, extending across a combination of two tracks that ascending and two tracks that descending. Two well-known processing methods, LiCSAR and DinSAR, were utilized to create the seismic deformation field and produce a time series of 2D deformation. One of the most powerful earthquakes to strike the region since 1900 happened on November 12, 2017, in Sarpol Zahab City located at the Iraq-Iran border. Another series of seismic events occurred approximately 100 kilometers to the south, near Mandali-Sumar in Iraq, on January 11, 2018. These earthquakes were noteworthy due to their intensity and proximity to populated areas, highlighting the potential risks and dangers associated with seismic activity in the region. This paper aims to create a map of neotectonics deformation of the ground along the Iraq-Iran borders from 2014 to 2023 and detect fault activity because of the earthquakes that occurred on November 12 and January 11, 2018, using inSAR techniques and confirmation on the processing, dissection, and applications of Sentinel-1 data backed by field, and tectonic information. The time series of 2D seismic deformation demonstrated a gradual decreasing pattern, which corresponded well with the distribution of aftershocks. The results show that the deformation for the Sarpol Zahab region extends to an area of 70 km2 and 90 km2, the greatest line-of-sight displacement of horizontal deformation is 20 cm and vertical displacement is 100 cm uplifting and 35 cm subsidence. The results of deformation for the Mandali-Sumar region extend to an area of 27 km2 and 23 km2, the greatest line of sight displacement of horizontal deformation is 5 cm and vertical displacement is 20 cm uplifting and 5 cm subsidence for ascending and descending data, respectively. The outcomes of this research will provide valuable insights into the investigation of tectonic events occurring in the region.
... According to [22,23], the phase of scatterers in SAR images is randomly distributed between 0~2 , and the real and imaginary parts of the composite echo are statistically independent zero-mean Gaussian random variables. The amplitude follows a Rayleigh distribution, and the intensity follows a negative exponential distribution. ...
... According to [22,23], the phase of scatterers in SAR images is randomly distributed between 0 ∼ 2π, and the real and imaginary parts of the composite echo are statistically independent zero-mean Gaussian random variables. The amplitude follows a Rayleigh distribution, and the intensity follows a negative exponential distribution. ...
The video synthetic aperture radar (ViSAR) system can utilize high-frame-rate scene motion target shadow information to achieve real-time monitoring of ground mobile targets. Modeling the characteristics of moving target shadows and analyzing shadow detection performance are of great theoretical and practical value for the optimization design and performance evaluation of ViSAR systems. Firstly, based on the formation mechanism and characteristics of video SAR moving target shadows, two types of shadow models based on critical size and shadow clutter ratio models are established. Secondly, for the analysis of moving target shadow detection performance in ViSAR systems, parameters such as the maximum detectable speed of moving targets, the minimum clutter backscatter coefficient, and the number of effective shadow pixels of moving targets are derived. Furthermore, the shadow characteristics of five typical airborne/spaceborne ViSAR systems are analyzed and compared. Finally, a set of simulation experiments on moving target shadow detection for the Hamas rocket launcher validates the correctness and effectiveness of the proposed models and methods.
... However, other than landslides, a number of unavoidable factors can disrupt and alter the surface backscattering mechanism in a SAR imagery, resulting in significant changes in the time-series of SAR parameters and false detection. These factors include changes in soil moisture, vegetation, rainfall, snowfall, anthropogenic activities, geometric layover and shadow (Lee et al., 1994;Czuchlewski et al., 2003;Oliver and Quegan, 2004;Mondini et al., 2021). The combined time-series of all SAR-derived parameters that are able to detect landslides may help tackle some of the problems related to false alarms in landslide detection arising from a single parameter time-series analysis (Czuchlewski et al., 2003;Shimada et al., 2014;Niu et al., 2021). ...
... The combination of all the SAR-derived parameters W. Wang et al. offers a comprehensive approach to overcome these challenges. (Lee et al., 1994;Czuchlewski et al., 2003;Oliver and Quegan, 2004;Shimada et al., 2014;Mondini et al., 2021). ...
... Acoustical measurements of the seafloor are known to show speckle modulations on top of the seafloor texture and backscattering strength. These speckle variations originate from the superimposed backscattering of randomly distributed scatterers within each resolution cell 8 . The statistics of speckle is well known, and will be adopted for our study where we approximate that neighbour samples are drawn from the same speckle distribution. ...
... Effect from layover of three surfaces with source levels of 0 dB, -5 dB and -10 dB, corresponding interferometric phases of 0 • , 120 • and -120 • , and white Gaussian noise of -20 dB. Left: The analytic PDF obtained from(8). Right: One realization of 100 samples drawn from the PDF (black), the individual contributions (red, green and blue) and contribution from noise (magenta). ...
... Finally, we have: In this article, the reflectivity R is assumed to follow a Gamma distribution, based on eq. 19 and 20, and in accordance with the Goodman model [24], [38]. In this case, the noise on the observations is a multiplicative speckle noise. ...
This article studies the use of airborne Global Navigation Satellite System Reflectometry (GNSS-R) techniques for remote sensing applications at regional scale. The objective is to classify the reflectivity of airborne GNSS signals in order to differentiate various reflective surfaces along the satellite traces. For this purpose, we propose a segmentation algorithm based on an online change point detector and an off-line change point localization estimate. Given the presence of speckle noise in GNSS signals, a homomorphic log-transformation is applied to mitigate this noise. In this context, the Cumulative Sum (CUSUM) change point detector and the Maximum Likelihood change point localization are designed for a Log-Gamma distribution. We show that the proposed radar segmentation system is able to automatically detect different land-forms along real flight experiments that took place in the Northern Region of France. Automatic classification using K-means clustering is applied to the segmented signals allowing to distinguish different segments of the signal.
... S YNTHETIC Aperture Radar (SAR) is one of the most advanced imaging technologies in the field of remote sensing. It has the characteristics of all-weather, all-day, and high-resolution imaging [1]. Polarimetric SAR (PolSAR) can simultaneously acquire echoes from four different polarization combinations, offering richer scattering information of ground objects compared to conventional SAR [2], [3]. ...
Polarimetric SAR (PolSAR) image classification is a critical application of remote sensing image interpretation. Most of the early algorithms that use hand-crafted features to divide the image into various scattering categories have a general classification performance. The Convolutional Neural Networks (CNNs) show superior performance in image processing with powerful nonlinear representation capabilities. However, they also require a large amount of labeled training data, which limits the practical use of CNNs in PolSAR image classification scene where labeled data are rare and expensive. To address the previous issue, this paper proposes a deep similarity clustering model with compound regularization for unsupervised PolSAR image classification. The proposed model combines an unsupervised feature extraction pipeline with Wishart distance metric and a deep clustering pipeline with feature similarity metric. The regularization combines two ingredients to preserve both the sharpness of edges and the semantic continuity of the image contents. The first is the differential constraint based on pixel-level features, and the second is the graph partition constraint based on superpixel-level features. Experiments prove the effectiveness of the proposed method on both spaceborne (RadarSat2) and airborne (E-SAR/UAVSAR) PolSAR images. The visual results and quantitative scores show that our method outperforms the traditional unsupervised methods and deep learning-based unsupervised methods with a large margin.
... Due to its unique imaging mechanism, SAR is capable of acquiring high-resolution images of the Earth's surface under any weather, climate, and time conditions (Li et al., 2014). The interpretation of SAR images is a challenging task that requires advanced algorithms and techniques to extract meaningful information from the data (Oliver and Quegan, 2004;Feng et al., 2021). SAR image target detection plays a crucial role in the interpretation process by identifying target classes and their spatial locations. ...
Due to the high mobility and strong concealment characteristics of synthetic aperture radar (SAR) targets, the corresponding SAR datasets exhibit few-shot data properties, and there is a significant lack of research on few-shot target detection methods in the SAR domain. Furthermore, this study is subject to the following limitations: the scarcity of SAR data and significant sample variations make it difficult to control class centers using existing methods, and the learned models tend to be biased towards base classes while easily confusing novel classes with base classes. These limitations hinder the generalization of knowledge from base classes when detecting novel class targets. In this work, we propose a novel few-shot SAR target detection method based on Gaussian meta-feature balanced aggregation (GMFBA), which is based on meta-learning. Specifically, we first propose two novel feature aggregation methods with Gaussian metrics, namely Gaussian projection distribution metric (GPDM) and Gaussian kernel mean embedding metric (GKMEM). By estimating class distribution with variational autoencoders to replace traditional class prototypes, we sample from robust distributions and measure projection Wasserstein distance and Gaussian kernel mean embedding distance with prior distributions, obtaining the best robust support features under the optimal measurement results. Then, based on GPDM and GKMEM, we propose a novel balanced inter-class uncorrelated aggregation (BICUA) method, which extracts support features of each class according to the proportion of samples and aggregates them with query features in a balanced manner, promoting feature representation between different classes and ensuring no interference between features to significantly reduce confusion between base classes and novel classes. Specifically, GMFBA outperforms the state-of-the-art method G-FSOD significantly in all settings, achieving state-of-the-art performance. In contrast, the novel class detection performance of GMFBA has shown an average improvement of 8.56% on split1 and split2 of the SRSDD-v1.0 dataset, and an average improvement of 1.41% on split1 and split2 of the MSAR-1.0 dataset. The code is available at https://github.com/Caltech-Z/GMFBA.
... Although this work offers insightful information about mapping flood extent in Himachal Pradesh with SAR data, it also recognises certain inherent limits of the SAR technology and our methodology. First off, there's a chance that SAR data misclassifies flood maps due to limitations in its ability to distinguish between flood waters and other wet surfaces like wetlands or damp soil [19]. Furthermore, Townsend [20] notes that dense vegetation can have a substantial impact on the radar signal, creating uncertainty when interpreting flood extents beneath wooded canopies. ...
One of the most destructive and frequent natural disasters in the world, flash floods cause millions of people to be displaced annually in addition to seriously harming livelihoods and infrastructure. It affects many ecological components and applications related to water management, natural resources, agriculture, human health, and economics. Himachal Pradesh saw an unprecedented amount of rainfall in June and July 2023, which resulted in exceptionally strong monsoon conditions from July 7 to July 10. Operating out of the Meteorological Centre in Shimla, the India Meteorological Department reported widespread, unusually heavy rainfall throughout the state during this time. Cloud cover often obstructs optical satellite data during the monsoon season, which has led to the investigation of alternate techniques for mapping floods. the whole satellite image processing process was carried out using Google Earth Engine (GEE). The Kangra District's 5739 km2 area was chosen as the study's area of interest. In these circumstances, traditional methods of mapping and monitoring flood-prone areas frequently fall short because of poor visibility and overcast skies during bad weather. Technology such as Synthetic Aperture Radar (SAR) becomes a ray of hope during these desperate times. When SAR data is combined with Google Earth Engine, it becomes even more user-friendly. Using SAR data and the powerful cloud processing platform Google Earth Engine (GEE), this study suggests a flood mapping technique. When it comes to mapping flood areas, the strength of SAR data and GEE surpasses boundaries and challenges. It is evidence of both technological advancement and human inventiveness. Such tools are more critical than ever as the world deals with the increasing effects of climate change and the rising frequency of extreme weather events.
... The likelihood takes into account the SSA model as well as the residual speckle noise after filtering. When SAR data are on dB scale, radiometric uncertainties are constant and data distribution can be assumed Gaussian [41], so the likelihood is normally distributed with mean (SSA(X) − (HH, HV)) and standard deviation given by the observation uncertainties summarized in Table 3. These uncertainties are below 10% (except for ID-16) and reflect the quality of the speckle filtering technique. ...
Lithium mining has become a controversial issue in the transition to green technologies due to the intervention in natural basins that impact the native flora and fauna in these environments. Large resources of this element are concentrated in Andean salt flats in South America, where extraction is much easier than in other geological configurations. The Pozuelos highland salt flat, located in northern Argentina (Salta’s Province), was chosen for this study due to the presence of different evaporitic crusts and its proven economic potential in lithium-rich brines. A comprehensive analysis of a 5.5-year-long time series of its microwave backscatter with Synthetic Aperture Radar (SAR) images yielded significant insights into the dynamics of their crusts. During a field campaign conducted near the acquisition of three SAR images (Sentinel-1, ALOS-2/PALSAR-2, and SAOCOM-1), field measurements were collected for computational modeling of the SAR response. The temporal backscattering coefficients for the crusts in the salt flat are directly linked to rainfall events, where changes in surface roughness, soil moisture, and water table depth represent the most critical variables. Field parameters were employed to model the backscattering response of the salt flat using the Small Slope Approximation (SSA) model. Salt concentration of the subsurface brine and the water table depth over the slightly to moderately roughed crusts were quantitatively derived from Bayesian inference of the ALOS-2/PALSAR-2 and SAOCOM-1 SAR backscattering coefficient data. The results demonstrated the potential for subsurface estimation with L-band dual-polarization images, constrained to crusts compatible with the feasibility range of the layered model.
... As mentioned previously, the Gamma and Weibull will also be analyzed in this paper. The Gamma distribution is used in the product model when the RCS is considered a constant [42]. The Gamma distribution can describe the characteristics of the RCS fluctuations of a heterogeneous terrain in high-resolution SAR images [43]. ...
We propose an algorithm based on Information Theory to detect changes in Ultra-Wideband (UWB) Very-High Frequency (VHF) Synthetic Aperture Radar (SAR) images with high performance and low complexity. Our algorithm models the clutter-plus-noise using six different distributions and computes a scalar statistic for each pixel based on a multi-temporal stack of images. With this statistic, it is then possible to apply hypothesis testing and classification methods to infer the occurrence of a change. In this context, we derive expressions necessary for the entropy-based statistics, including the entropy variance for the Weibull and Rice distributions. We also evaluate the computational time complexity of the algorithm for each distribution studied. Furthermore, a masking strategy is used to reduce false alarms significantly. We show that the mask mapping assumptions are mild in scenarios with stacks of images, allowing its use in many scenarios. Our algorithm achieves a false alarm rate (FAR) of 0.08 and a probability of detection (PD) of 100%, outperforming existing methods on the CARABAS II data set.
... They are limited to interval 1 ≤ n ≤ N . The normalized correlation of x(n) with s(n) is defined as where x(n) is considered to conform to the Rayleigh distribution, i.e., it has the probability density function This is derived from the assumption that the real part and the imaginary part of the complex range profile are independent Gaussian random variables with the mean 0 and the standard deviation σ [24]. It can be shown that the mean, the mean square, and the variance of x(n) are, respectively, The norm of x(n), ε , can be expressed as where x(n) is assumed to be ergodic. ...
In this paper, a novel algorithm is presented for warhead recognition in the defense of ballistic missiles. The range profiles from the warheads of interest in typical illumination directions form a dataset. First, each range profile in the dataset is compared to the range profile of the target under observation, and the most similar range profile is found. Then, the observed target is considered as a warhead if the deviation of its range profile from the most similar range profile is less than or equal to a threshold. The threshold is chosen such that the detection rate is a constant. The simulation results verify the effectiveness of the proposed algorithm. Since the threshold is automatically calculated according to the detection rate, this algorithm has a larger applicability than the current methods based on range-profile matching.
... The detection of icebergs using SAR data often employs conventional constant false alarm rate (CFAR) using a sliding window. Targets are discriminated by looking at anomalies in the backscattering when comparing a target window with a clutter window [17]. The threshold is set using statistical tests and any target brighter than the threshold triggers a detection. ...
Satellite monitoring of icebergs in the Arctic region is paramount for the safety of shipping and maritime activities. The potential of PolSAR data in enhancing detection capabilities of icebergs under interchangeable and challenging conditions is explored in this work. We introduce RADARSAT-2 (RS2) quad-pol C-band data to detect icebergs in Kongsfjorden, Svalbard. The location contains two tidewater glaciers and is chosen because multiple processes are present in this region such as ice formation and its relationship with the glaciers, freshwater discharge. Six state-of-the-art detectors are tested for detection performance. These are the dual intensity polarisation ratio anomaly detector (iDPolRAD), polarimetric notch filter (PNF), polarimetric match filter (PMF), symmetry, polarimetric whitening filter (PWF), optimal polarimetric detector (OPD). Additionally, we also tested the parameters of the Cloude-Pottier decomposition. In this study, we make use of a ground-based radar for validation and comparison with satellite images. We show that in calm sea-state conditions, the OPD and PWF detectors give high Probability of Detection (PD) values of 0.7-0.8 when the Probability of False Alarm (PF) value is 0.01-0.05, compared to choppy sea conditions where the same detectors have degraded performance (PD = 0.5-0.7). Target to clutter ratio (TCR) values for each polarization channel is also extracted and compared to the icebergs’ dimensions. The ground-based radar shows higher values in TCR, compared to satellite images. These findings corroborate previous work and show that sea ice activity, surface roughness, incidence angle, weather and sea state conditions all affect the sensitivity of the detectors for this task.
... Analyzing the noise properties of the SAR images used, we observed that the real and imaginary part of the images have an approximate Gaussian-shaped distribution (cf. [25]). We measured the statistics of this Gaussian noise in homogeneous and low amplitude areas of the reference image. ...
Image registration is a crucial step in interferometric synthetic aperature radar (InSAR) and in SAR tomography. Generally, a displacement of the sensors, e.g. due to different flight tracks, causes an image distortion that is dependent on the terrain height that is being observed. While ground truth digital elevation map (DEM) data are often available to roughly compensate this distortion, there are application scenarios where the acquisition paths are too irregular or DEMs are not available. In such cases, image registration via image processing is a suitable choice. While the SAR community prefers patch-based correlation techniques, the computer vision community has investigated the same issue from technically different points of view. In this article, we study the performance of correlation-and computer vision-based image registration techniques for the image registration problem in SAR, in particular in airborne SAR without ground truth. We show that computer vision algorithms can outperform correlation-based techniques.
... This work focuses on color image denoising in the presence of multiplicative noise. Multiplicative noise usually occurs under coherent image systems such as synthetic aperture radar (SAR) [1], ultrasound imaging [2], and laser imaging [3]. Owing to the coherent nature of these image acquisition procedures, most of the information in an original image may be lost when corrupted by multiplicative noise. ...
In this article, we propose a novel group-based sparse representation (GSR) model for restoring color images in the presence of multiplicative noise. This model consists of a convex data-fidelity term, and two regularizations including GSR and saturation-value-based total variation (SVTV). The data-fidelity term is suitable for handling heavy multiplicative noise. GSR enables the retention of textures and details while sufficiently removing noise in smooth regions without producing the staircase artifacts engendered by total variation-based models. Furthermore, we introduce a multi-color channel-based GSR that involves coupling between three color channels. This avoids the generation of color artifacts caused by decoupled color channel-based methods. SVTV further improves the visual quality of restored images by diminishing certain artifacts induced by patch-based methods. To solve the proposed nonconvex model and its subproblem, we exploit the alternating direction method of multipliers, which contributes to an efficient iterative algorithm. Numerical results demonstrate the outstanding performance of the proposed model compared to other existing models regarding visual aspect and image quality evaluation values.
... The constructed scenario spans 1024 × 1024 pixels, featuring a distributed target covering an area of 101 × 101 pixels at its core. In line with Oliver and Quegan's research [26], an equivalent phase center was assigned to every pixel within the target area. The amplitude of each pixel adheres to a Rayleigh distribution that is independent and identically distributed (with a mean µ = πσ 0 2 and a variance σ 2 = p(1− π 4 )σ 0 2 , where σ 0 is the backscattering coefficient). ...
The integration of deep neural networks into sparse Synthetic Aperture Radar (SAR) imaging has been explored to enhance SAR imaging performance and reduce the system’s sampling rate. However, the scarcity of training samples and mismatches between the training data and the SAR system pose significant challenges to the method’s further development. In this paper, we propose a novel SAR imaging approach based on Deep Image Prior Powered by RED (DeepRED), enabling unsupervised SAR imaging without the need for additional training data. Initially, DeepRED is introduced as the regularization technique within the sparse SAR imaging model. Subsequently, variable splitting and the Alternating Direction Method of Multipliers (ADMM) are employed to solve the imaging model, alternately updating the magnitude and phase of the SAR image. Additionally, the SAR echo simulation operator is utilized as an observation model to enhance computational efficiency. Through simulations and real data experiments, we demonstrate that our method maintains imaging quality and system downsampling rate on par with deep neural network-based sparse SAR imaging, but without the requirement for training data.
... Doppler shift is a measure of the phase shift of an FM signal. Aperture system-uses a combination of signal processing methods to achieve resolution in the direction of the path [15,16]. ...
Synthetic Aperture Radar (SAR) was developed to map the terrain without the use of large antennas. Combined array is one of the radar methods that is applied from an aircraft or a space platform and in which an effective aperture of the antenna is created in a combined manner. The images obtained by these radars are very accurate. On the other hand, the surface of the earth always changes due to various factors. Accurate identification of these changes can be used in many applications, especially in Iraq due to its semidesert structure. In this research, an improved surface change detection algorithm based on morphological transformation, two-stage center-constrained FCM algorithm (TCCFCM) clustering and deep learning is presented. The simulation of the method of this research on MATLAB software shows that the proposed method is fast and inexpensive. The accuracy was 99.7%.
... The constructed scenario spans 1024 × 1024 pixels, featuring a distributed target covering an area of 101 × 101 pixels at its core. In line with Oliver and Quegan's research [29], an equivalent phase center was assigned to every pixel within the target area. The amplitude of each pixel adheres to a Rayleigh distribution that is independent and identically distributed (with a mean µ = πσ 0 2 and a variance σ 2 = (1 − π 4 )σ 0 , where σ 0 is the backscattering coefficient). ...
The integration of deep neural networks into sparse synthetic aperture radar (SAR) imaging is explored to enhance SAR imaging performance and reduce the system’s sampling rate. However, the scarcity of training samples and mismatches between the training data and the SAR system pose significant challenges to the method’s further development. In this paper, we propose a novel SAR imaging approach based on deep image prior powered by RED (DeepRED), enabling unsupervised SAR imaging without the need for additional training data. Initially, DeepRED is introduced as the regularization technique within the sparse SAR imaging model. Subsequently, variable splitting and the alternating direction method of multipliers (ADMM) are employed to solve the imaging model, alternately updating the magnitude and phase of the SAR image. Additionally, the SAR echo simulation operator is utilized as an observation model to enhance computational efficiency. Through simulations and real data experiments, we demonstrate that our method maintains imaging quality and system downsampling rate on par with deep-neural-network-based sparse SAR imaging but without the requirement for training data.
... Due to its ability to operate independently of atmospheric and sunlight conditions, synthetic aperture radar (SAR) offers advantages over optical remote sensing systems. Automatic target recognition (ATR) is a crucial application of SAR systems, traditional techniques relied on handcrafted features such as the shape, size, and intensity of objects in the images (Oliver and Quegan, 2004). However, these techniques faced limitations as they required manual feature extraction and were susceptible to variations in conditions, object orientations, and configurations Wu et al. (2023a) and Yuan et al. (2023). ...
Introduction
Decades of research have been dedicated to overcoming the obstacles inherent in synthetic aperture radar (SAR) automatic target recognition (ATR). The rise of deep learning technologies has brought a wave of new possibilities, demonstrating significant progress in the field. However, challenges like the susceptibility of SAR images to noise, the requirement for large-scale training datasets, and the often protracted duration of model training still persist.
Methods
This paper introduces a novel data augmentation strategy to address these issues. Our method involves the intentional addition and subsequent removal of speckle noise to artificially enlarge the scope of training data through noise perturbation. Furthermore, we propose a modified network architecture named weighted ResNet, which incorporates residual strain controls for enhanced performance. This network is designed to be computationally efficient and to minimize the amount of training data required.
Results
Through rigorous experimental analysis, our research confirms that the proposed data augmentation method, when used in conjunction with the weighted ResNet model, significantly reduces the time needed for training. It also improves the SAR ATR capabilities.
Discussion
Compared to existing models and methods tested, the combination of our data augmentation scheme and the weighted ResNet framework achieves higher computational efficiency and better recognition accuracy in SAR ATR applications. This suggests that our approach could be a valuable advancement in the field of SAR image analysis.
... , is often added to represent the phase of ab S′ [28]. Once the ab S′ is obtained, we can use (5) and (6) to get the scattering coefficients in the globe coordinate system SPM ab γ . ...
The two-scale method (TSM) is an effective approach to simulate electromagnetic (EM) scattering from rough sea surface, to interpret physical interaction of EM wave with rough sea surfaces. The sea surface roughness spectrum and the cutoff wavenumber are two key factors affecting the accuracy of TSM simulation. This paper takes the geophysical model functions (GMFs) of different band as reference, and quantitatively studies the two factors based on the numerical simulation of different polarization backscattering coefficients. The numerical simulation demonstrates that Apel spectrum is more suitable for the backscattering simulation in C/X/Ku-band, and the optimal cutoff wavenumbers are also derived for TSM simulation of EM scattering in C/X/Ku-band. Then an empirical model function of optimal cutoff wavenumbers with respect to frequency, incident angle and wind speed are obtained with regression method. Finally, the simulation results with the new model function of cutoff wavenumber are compared with the GMF and measured data in different bands and they are all consistent. Moreover, a physical explanation for the optimal cutoff wavenumber is also given.
... Assuming that the single-look complex image of the radar satisfies the complex circular Gaussian distribution [23], the amplitude of the temporal information A(P) = [A 1 , A 2 , L, A N ] of any pixel P satisfies the Rayleigh distribution: ...
Statistically homogeneous pixel (SHP) selection is an important process in the distributed scatterer interferometric synthetic aperture radar (DS-InSAR) approach. However, prevalent methods struggle to appropriately balance the efficiency and accuracy of selection. To solve this problem, the authors of this study improved the Hypothesis Test of Confidence Interval (HTCI) to propose an adaptive method to select the level of saliency and confidence interval for the HTCI, called Adp-HTCI. The proposed method can accurately select homogeneous pixels while inheriting the high efficiency of the HTCI. Once homogeneous pixels have been chosen, the eigenvalue decomposition of the covariance matrix is used to optimize their phase and perform temporal processing. We used the proposed method along with data on 31 scenes from the Sentinel-1 satellite from 2 June 2021 to 28 May 2022 to monitor the deformation of the surface of the fire zone in the Sikeshu coalfield. The selection results of homogeneous pixels indicate that the proposed Adp-HTCI algorithm can increase the number of selected homogeneous pixels while ensuring the accuracy of the selection results, thereby enhancing the estimation accuracy and reliability of subsequent parameter solving. The DS-InSAR results showed that the cumulative maximum subsidence in the study area within a year reached—138 mm and the point density used by the DS-InSAR approach was 17.28 times higher than that used by the StaMPS approach. The results of cross-analysis with the results of StaMPS verified the accuracy of the DS-InSAR-based approach.
... SAR image interpretation is very difficult for several reasons. The unique geometric characteristics of SAR images increase the difficulty interpretation [11]; the inherent coherent speckle noise of SAR images causes target edges to be blurred, and the clarity to decrease, requiring completely different methods for SAR image interpretation. SAR images also have multiple reflection effects, false phenomena, doppler frequency shifts, etc. ...
Synthetic Aperture Radar (SAR) to electro-optical (EO) image translation is a fundamental task in remote sensing that can enrich the dataset by fusing information from different sources. Recently, many methods have been proposed to tackle this task, but they are still difficult to complete the conversion from low-resolution images to high-resolution images. Thus, we propose a framework, SAR2EO, aiming at addressing this challenge. Firstly, to generate high-quality EO images, we adopt the coarse-to-fine generator, multi-scale discriminators, and improved adversarial loss in the pix2pixHD model to increase the synthesis quality. Secondly, we introduce a denoising module to remove the noise in SAR images, which helps to suppress the noise while preserving the structural information of the images. To validate the effectiveness of the proposed framework, we conduct experiments on the dataset of the Multi-modal Aerial View Imagery Challenge (MAVIC), which consists of large-scale SAR and EO image pairs. The experimental results demonstrate the superiority of our proposed framework, and we win the first place in the MAVIC held in CVPR PBVS 2023.
... Studies indicate that noise in an image can reduce classification accuracy by up to 10% (Chen et al. 2016). Multiplicative and additive noises or blurring induced by the radar's imaging direction changes typically cause SAR image failure (Oliver and Quegan 1998;Yin et al. 2015). Angle shifts and refractions can change the object's appearance and its scattering pattern (Chang and You 2018). ...
Identifying a target accurately in the presence of noise in synthetic aperture radar (SAR) images poses significant challenges considering various parameters, such as viewing angle and configuration changes. In this paper, SAR images are classified using the proposed convolutional neural network. The limitation of data concerning moving and stationary target acquisition and recognition (MSTAR) database led us to develop a primary network (learning network) by employing the proposed N-Sigmoid function and training it using eight different satellite databases with images more than MSTAR. We saved the best weights for the main network. Next, we replace the weights of the two fully connected (FC) layers of the ResNet-50 with the saved weights of the training network and employed the main network for classification. Our proposed method’s primary contribution is twofold, based on the availability of additional information in heterogeneous SAR images and differences in intensity and clarity. First, we utilize the weights learned by the training network, and second, we keep negative neurons and neurons close to zero active instead of deactivating them to facilitate better training of the main network. Our proposed method outperformed other references in terms of sensitivity in all classes and achieved an accuracy of 98% for classification with good machine weight learning.
... It deviates from the traditional assumption of additive noise and follows a different statistical distribution, such as the Gamma or Rayleigh distribution. Multiplicative noise can arise in various imaging scenarios, such as synthetic aperture radar (SAR) [1], ultrasound [2], laser images [3], and other coherent image systems [4]. The mathematical model for image degradation caused by blur and multiplicative noise can be expressed as: ...
In this paper, we propose a novel convex variational model for image restoration with multiplicative noise. To preserve the edges in the restored image, our model incorporates a total variation regularizer. Additionally, we impose an equality constraint on the data fidelity term, which simplifies the model selection process and promotes sparsity in the solution. We adopt the alternating direction method of multipliers (ADMM) method to solve the model efficiently. To validate the effectiveness of our model, we conduct numerical experiments on both real and synthetic noise images, and compare its performance with existing methods. The experimental results demonstrate the superiority of our model in terms of PSNR and visual quality.
... In the simulation experiments, we generate a 128 × 128 pixels scene with distributed targets that occupy 50 × 50 pixels each and point targets in the blank, as Fig. 3(a) shows. According to [36], the complex distributed targets should be Rayleigh distributed in amplitude and uniformly distributed in phase for simulation, as Fig. 3(b) shows. A 1024 × 1024 Fourier matrix is chosen as the measurement matrix. ...
Composite regularization models are widely used in sparse signal processing, making multiple regularization pa-rameters selection a significant problem to be solved. Variety kinds of composite regularization models are used in sparse microwave imaging, including ℓ
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub>
and ℓ
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub>
penalty, nonconvex and TV penalty, combined dictionary, etc. In this article, a new adaptive multiple regularization parameters selection method named L-hypersurface is proposed. The effectiveness of the pro-posed method is verified by experiments. Simulation experiments indicate that the selected optimal regularization parameters have satisfied reconstruction results, both visually and numerically. Furthermore, experiments on Gaofen-3 SAR satellite data are also exploited to show the performance of the proposed method.
... As the crop grows the canopy structure changes increasing both backscattering and the randomness of scattering (McNairn et al., 2018). The plant components also attenuate the microwave signal depending on the polarization and frequency employed (Fleischman et al., 1996;Oliver and Quegan, 1997). The SAR response of C-band sensor is related to crop structure, biomass development, soil condition and has led to many studies for crop monitoring. ...
In the framework of the comparison of Synthetic Aperture Radar (SAR) imagery from the Magellan space mission and the VISAR and VenSAR radar instruments which will be onboard the forthcoming VERITAS and EnVision missions to Venus, the problem of the disparity between the resolutions of the images arises when attempting to define a test statistic with which to detect changes. Reliable change detection requires equivalent spatial resolutions which, for the two different images, inevitably involves different equivalent number of looks after speckle-reduction processing. This study presents a method to address this scenario using a Generalized Beta Prime Distribution as a probability density function (
PDF
) which is fit to the histogram of the ratio between the two intensity images. The work demonstrates and verifies the properties of the function, highlights its most useful traits, and elaborates on the mathematical procedure required to achieve a meaningful change detection in line with the classic theory of equal number of looks. The results show that the method accurately describes the ratio histogram of two SAR intensity images with different number of looks. Furthermore, they demonstrate the adaptability of the method to the presence of high pixel correlation between the images, and validate its robustness in the presence of textural complexity when the texture patterns of the images are similar.
Sparse synthetic aperture radar (SAR) imaging has emerged as a reliable microwave imaging scheme in the recent decade and excels in down-sampling reconstruction and full-sampling performance improvements such as noise, sidelobe, speckle, and ambiguity suppression. To utilize complex image products of sparse reconstruction for improvement in polarimetric, interferometric, and tomographic SAR imaging, it is necessary to evaluate the phase preservation of sparse SAR imaging. In this study, we first introduce the general alternating direction method of multipliers (ADMM) as the universal framework for sparse reconstruction algorithms and adopt chirp scaling algorithm (CSA)-based azimuth-range decouple operators to avoid expensive data storage and processing. Further, we theoretically analyze the phase preservation of the sparse reconstruction algorithm through a comparison with the reconstruction results of CSA. Finally, we conduct the interferometric offset test on the sparse reconstruction results of simulated and real Gaofen-3 (GF-3) SAR data, demonstrating the phase-preserving ability of sparse methods.
Despeckling is a crucial noise reduction task in improving the quality of synthetic aperture radar (SAR) images. Directly obtaining noise-free SAR images is a challenging task that has hindered the development of accurate despeckling algorithms. The advent of deep learning has facilitated the study of denoising models that learn from only noisy SAR images. However, existing methods deal solely with single-polarization images and cannot handle the multi-polarization images captured by modern satellites. In this work, we present an extension of the existing model for generating single-polarization SAR images to handle multi-polarization SAR images. Specifically, we propose a novel self-supervised despeckling approach called channel masking, which exploits the relationship between polarizations. Additionally, we utilize a spatial masking method that addresses pixel-to-pixel correlations to further enhance the performance of our approach. By effectively incorporating multiple polarization information, our method surpasses current state-of-the-art methods in quantitative evaluation in both synthetic and real-world scenarios.
Due to the shortcomings of traditional interferometric synthetic aperture radar (InSAR) technologies in mining subsidence monitoring, such as a low density of monitoring points and difficulty obtaining fine surface deformation information, this paper proposes a mining area surface deformation monitoring method based on distributed scatterers InSAR while improving the phase information optimization strategy. This method obtains homogeneous points using the hypothesis test of confidence interval algorithm and constructs an adaptive phase optimization method based on the Goldstein principal phase filtering and Fisher information matrix weighting. It effectively preserves the information of deformation fringes, particularly in regions with dense interferometric fringes, and obtains detailed deformation information from the study area through time processing. In the experiments, 63 Sentinel-1 images were used to extract surface subsidence information for the Peibei mining area from September 24, 2018, to November 12, 2020. Compared with the Permanent Scatterers InSAR (PS-InSAR) results, the point density increased by a factor of 4.2. The correlation coefficient between the homonymous points obtained with the two methods and the deformation rate is 0.94, indicating that they have a good consistency. The monitoring results show that the six mining areas in Peibei have different degrees of subsidence during the monitoring period with a clear nonlinear trend, and the maximum cumulative subsidence over time exceeds 350 mm. The analysis shows that the improved DS-InSAR monitoring results are in line with the general law of mining subsidence and have practical application value.
The paper proposes a generalized optimal sensing principle for synthetic aperture radar (SAR) imaging, maximizing the mutual information between the sensed object and the reconstructed image with the optimal SAR measurement matrix. Inspired by Shannon’s capacity theorem, the SAR sensing capacity is derived, which represents the maximum mutual information that can be acquired per unit distance or unit area in 1-D or 2-D scenarios. The SAR sensing capacity serves as a theoretical performance bound, guiding the design of SAR sensing systems and enabling reasonable estimation of systems’ performance. Additionally, this paper analyzes the relationship between system parameters and the column correlation of SAR measurement matrices, guiding the design of the SAR system. Furthermore, the optimal sensing principle is applied to the variable-resolution SAR (VR-SAR) imaging system. Theoretical simulations are conducted to verify the feasibility of the optimal sensing principle, and examine the relationship between column correlation and SAR parameters. Additionally, the advantages of VR-SAR based on the optimal sensing principle are compared with those of the conventional strip-map SAR mode.
Urban infrastructure is an important part of supporting the daily operation of a city. The stability of infrastructure is subject to various deformations related to disasters, engineering activities, and loadings. Regular monitoring of such deformations is critical to identify potential risks to infrastructure and take timely remedial actions. Among the advanced geodetic technologies available, radar interferometry has been widely used for infrastructure stability monitoring due to its extensive coverage, high spatial resolution, and accurate deformation measurements. Specifically, spaceborne InSAR and ground-based radar interferometry have become increasingly utilized in this field. This paper presents a comprehensive review of both technologies for monitoring urban infrastructures. The review begins by introducing the principles and their technical development. Then, a bibliometric analysis and the major advancements and applications of urban infrastructure monitoring are introduced. Finally, the paper identifies several challenges associated with those two radar interferometry technologies for monitoring urban infrastructure. These challenges include the inconsistent in the distribution of selected measurements from different methods, obstacles arising from rapid urbanization and geometric distortion, specialized monitoring techniques for distinct urban features, long-term deformation monitoring, and accurate interpretation of deformation. It is important to carry out further research to tackle these challenges effectively.
Multi-channel high-resolution and wide-swath (MC-HRWS) synthetic aperture radar (SAR) systems with digital beamforming (DBF) in azimuth can suppress doppler ambiguity to realize HRWS SAR imaging. The imaging position deviation of the moving ship in SAR image is caused by the radial velocity, which can be estimated to accurately positioning the moving ship in SAR image. For MC-HRWS SAR system, radial velocity will cause Doppler spectrum distortion and azimuth ambiguous, which will reduce the accuracy of traditional single-channel SAR radial velocity estimation method. This paper presents a high accuracy radial velocity estimation approach for ships based on maximum likelihood (ML) in MC-HRWS SAR system. The influence of the sea background is eliminated by estimating its Doppler centroid, and extracting the azimuth antenna pattern (AAP) and the system noise power from the signal of background sea clutter. In contrast to the conventional radial velocity estimation methods based on echo data for MC-HRWS SAR, the proposed method enables direct estimation based on focused SAR single-look complex (SLC) data devoid of auxiliary information, the sea clutter interference is small, and the interaction of multiple ships under the same beam irradiation is avoided. The effectiveness of the method is demonstrated via simulation and GaoFen-3 (GF-3) data of ultra-fine strip-map (UFS) mode. Furthermore, the obtained results show that the estimated radial velocity is consistent (6% root-mean-square deviation) with that provided velocity information of AIS data.
The Black Sea as inland water body is the main trade route of the countries in the region. With increased trade volume and the region's coverage of the world's fifth most important oil transfer transition, it has been a region where oil spills have been seen continuously over the years.Within the scope of the study, oil spill detection and analysis were carried out on Sentinel-1 radar images obtained between 2015 and 2019. For the purpose of speckle noise reduction, Local Median, Frost, Gamma MAP and Lee Sigma filters were performed. The best despeckling filter was determined and it was used for detection and morphological analysis. Based on the results of the different images, the most inaccurate method was observed as Canny Edge detection, while Otsu Thresholding was the method with the best reliability within itself. .The outcomes of clustering algorithms were found to be similar to each other. Also, the same outcomes were shown for both classification algorithms.
ResearchGate has not been able to resolve any references for this publication.