Synthetic Aperture Radar Classification Research Articles

Region-Based Classification of SAR Images Using Kullback–Leibler Distance Between Generalized Gamma Distributions

For the classification of synthetic aperture radar (SAR) images, traditional pixel-based Bayesian classifiers suffer from an intrinsic flaw that categories with serious overlapped probability density functions cannot be well classified. To solve this problem, in this letter, a region-based classifier for SAR images is proposed, where regions, instead of individual pixels, are treated as elements for classification. In the algorithm, each region is assigned to the class that minimizes a criterion referring to the Kullback–Leibler distance. Besides, the generalized gamma distribution $(\mathrm{G}\Gamma\mathrm{D} ) $ , a flexible empirical model, is employed for the statistical modeling of SAR images. Finally, with a synthetic image and an actual SAR image acquired by the EMISAR system, the effectiveness of the proposed algorithm is validated, compared with the pixel-based maximum-likelihood method and two region-based Bayesian classifiers.

Particle Filter Sample Texton Feature for SAR Image Classification

This letter presents a novel approach to learning the correct sampling positions by introducing a particle filter. A filtering, labeling, and statistics framework we previously proposed is applied to construct a complete texture descriptor named particle filter sample texton (PFST) feature for the classification of synthetic aperture radar (SAR) images. First, the gray values of the key points tracked by a particle filter in the local image patch are concatenated into a vector. Second, the vectors are labeled using a texton dictionary clustered from the training images. Finally, the histogram statistics is performed on these labels to generate the feature vectors for classification. The proposed method is more robust in terms of speckle noise and extremely low signal-to-noise ratio than those of the existing fixed-point and random sampling methods that play a significant role in popular binary textural descriptors. The experiments conducted on the TerraSAR image present evidence that the key points tracked by the particle filter effectively preserve the texture information, and the PFST feature performs best in the extreme situations.

Histogram-Based Contextual Classification of SAR Images

We propose a spatially dependent mixture model for contextual classification of synthetic aperture radar (SAR) images. The proposed mixture model is based on the local image histograms modeled by multinomial densities. The contextual information is included into the mixture model both in the pixel and the class label domain by using local histograms and autologistic regression, respectively. Based on the classification results obtained on real TerraSAR-X images, it is shown that the proposed model is capable of more accurately classifying the pixels particularly in the heterogeneous regions, such as urban areas, compared with the conventional mixture model.

The Jeffries–Matusita distance for the case of complex Wishart distribution as a separability criterion for fully polarimetric SAR data

In multidimensional observations, many classification algorithms (supervised or unsupervised) require the selection of optimum bands in which the classes are most distinct. The Jeffries–Matusita (JM) distance is widely used as a separability criterion for optimal band selection and evaluation of classification results. Its original form is based on the assumption of normal distribution of the data. However, in the case of the covariance/coherency matrix of synthetic aperture radar (SAR) polarimetry, the data follow the complex Wishart distribution. In this article, we calculate the JM separability criterion for the case of the complex Wishart distribution. The updated formulation is used for: (1) the estimation of the separability between classes in fully polarimetric SAR data and to evaluate two standard polarimetric SAR classification algorithms, the Wishart and the expectation maximization algorithms, and (2) the classification of fully polarimetric SAR images based on the derived JM separability for the case of complex Wishart distribution. Fully polarimetric RADARSAT-2 images over sea ice in the Canadian Arctic are used to classify different ice surfaces and open water.

Unsupervised polarimetric synthetic aperture radar classification of large-scale landslides caused by Wenchuan earthquake in hue-saturation-intensity color space

A simple and effective approach for unsupervised classification of large-scale landslides caused by the Wenchuan earthquake is developed. The data sets used were obtained by a high-resolution fully polarimetric airborne synthetic aperture radar system working at X-band. In the proposed approach, Pauli decomposition false-color RGB imagery is first transformed to the hue-saturation-intensity (HSI) color space. Then, a good combination of k-means clustering and HSI imagery in different channels is used stage-by-stage for automatic landslides extraction. Two typical case studies are presented to evaluate the feasibility of the proposed scheme. Our approach is an important contribution to the rapid assessment of landslide hazards.

New SAR Algorithm Based on Orthogonal Projections for MMT Detection and Interference Reduction

We develop a new synthetic aperture radar (SAR) algorithm based on physical models for the detection of a man-made target (MMT) embedded in strong interferences (trunks of a forest). These physical models for the MMT and the interferences are integrated in low-rank subspaces and are based on scattering and polarimetric properties. Several images, called subspace SAR images, can be generated and combined considering these subspace models. We then propose a new approach for target detection and interference reduction based on the combination of SAR subspace images. We show that our SAR algorithm outperforms the classical SAR imagery algorithm on both simulated data and real data in the context of foliage penetration detection.

Piecewise and Wiener Filter-Based SAR Techniques for Monostatic Microwave Imaging of Layered Structures

In this paper, two new techniques for microwave imaging of layered structures are introduced. These techniques were developed to address the limiting issues associated with classical synthetic aperture radar (SAR) imaging techniques in generating focused and properly-positioned images of embedded objects in generally layered dielectric structures. The first method, referred to as piecewise SAR (PW-SAR), is a natural extension of the classical SAR technique, and considers physical and electrical properties of each individual layer and the discontinuity among them. Although this method works well with low loss dielectric media, its applicability to lossy media is limited. This is due to the fact that this method does not consider signal attenuation. Moreover, multiple reflections within each layer are not incorporated. To improve imaging performance in which these important phenomena are included, a second method was developed that utilizes the Green's function of the layered structure and casts the imaging approach into a deconvolution procedure. Subsequently, a Wiener filter-based deconvolution technique is used to solve the problem. The technique is referred to as Wiener filter-based layered SAR (WL-SAR). The performance and efficacy of these SAR based imaging techniques are demonstrated using simulations and corresponding measurements of several different layered media.

ROC‐based calibration of flood inundation models

AbstractThe use of spatial patterns of flood inundation (often obtained from remotely sensed imagery) to calibrate flood inundation models has been widespread over the last 15 years. Model calibration is most often achieved by employing one or even several performance measures derived from the well‐known confusion matrix based on a binary classification of flooding. However, relatively early on, it has been recognized that the use of commonly reported performance measures for calibrating flood inundation models (such as the F measure) is hampered because the calibration procedure commonly utilizes only one possible solution of a wet/dry classification of a remote sensing image [most often acquired by a synthetic aperture radar (SAR)] to calibrate or validate models and are biased towards either over‐prediction or under‐prediction of flooding. Despite the call in several studies for an alternative statistic, to this date, very few, if any, unbiased performance measure based on the confusion matrix has been proposed for flood model calibration/validation studies. In this paper, we employ a robust statistical measure that operates in the receiver operating characteristics (ROC) space and allows automated model calibration with high identifiability of the best model parameter set but without the need of a classification of the SAR image. The ROC‐based method for flood model calibration is demonstrated using two different flood event test cases with flood models of varying degree of complexity and boundary conditions with varying degree of accuracy. Verification of the calibration results and optional SAR classification is successfully performed with independent observations of the events. We believe that this proposed alternative approach to flood model calibration using spatial patterns of flood inundation should be employed instead of performance measures commonly used in conjunction with a binary flood map. © 2013 California Institute of Technology. Hydrological Processes © 2013 John Wiley & Sons, Ltd.

SUB-APERTURE FOCUSING ALGORITHM OF GEOSYNCHRONOUS SAR

Abstract. Synthetic Aperture Radar (SAR) on the geosynchronous orbit has the advantage in remote sensing, with high resolution to provide a large coverage in a short visit time. the space-variant coverage and range-azimuth couple of GEO SAR are required to compensate, using the electrical beam steering, the irregular data acquisition can be convert to classical SAR modes, such as stripmap, spotlight, sliding spotlight, and terrain observation by progressive scans (TOPS) SAR, the dopper analysis of these work modes of SAR have been studied. Based on deramping processing and improved subaperture chirp scaling algorithm, a novel focusing approach is presented in this paper, whose performance is determined by the work mode, data acquisition and the deramping processing. Data of each mode can be focused by utilizing this presented approach with the information of subapterture centroid time and range rotation center. Simulation results are presented to validate the analysis of the proposed approach.

Unsupervised classification of SAR images using normalized gamma process mixtures

We propose an image prior for the model-based nonparametric classification of synthetic aperture radar (SAR) images that allows working with infinite number of mixture components. In order to enclose the spatial interactions of the pixel labels, the prior is derived by incorporating a conditional multinomial auto-logistic random field into the Normalized Gamma Process prior. In this way, we obtain an image classification prior that is free from the limitation on the number of classes and includes the smoothing constraint into classification problem. In this model, we introduced a hyper-parameter that can control the preservation of the important classes and the extinction of the weak ones. The recall rates reported on the synthetic and the real TerraSAR-X images show that the proposed model is capable of accurately classifying the pixels. Unlike the existing methods, it applies a simple iterative update scheme without performing a hierarchical clustering strategy. We demonstrate that the estimation accuracy of the proposed method in number of classes outperforms the conventional finite mixture models.

Local Primitive Pattern for the Classification of SAR Images

This paper proposes a new method for the classification of synthetic aperture radar (SAR) images based on a novel feature vector. The method aims at combining the intensity information of pixels with spatial information and structural relationships. Unlike classical approaches which define a static neighborhood via a rectangular moving window of predefined size and relate spatial information for each center pixel to all the pixels within that window, the local primitives (LPs) proposed in this study provide us with an adaptive neighborhood so that spatial information for each center pixel is extracted only from the related pixels in its neighborhood. LPs correspond to local homogeneous connected components that describe the pixel neighborhood more consistently than the fixed size window approach. A feature vector, called as the LP pattern (LPP), is constructed for each pixel. The feature vector includes information about the sizes, intensity levels, and contrast differences of LPs within a disk whose center is the pixel under consideration as well as the repetitive frequency of LPs outside that disk. Finally, a kernel-based support vector machine is used with the proposed feature vectors for the classification of SAR images. Experimental analysis presents that the new feature extraction technique is well suited to depict spatial information and structural relationships and it yields promising results for the classification of SAR images when compared to common features such as gray-level co-occurrence matrix and Gabor coefficients.

Radiofrequency interference suppression in synthetic aperture radar based on singular spectrum analysis with extended – FAPI subspace tracking

To obtain sufficient resolution, the operating band of synthetic aperture radar (SAR) must extend over several tens of megahertz. Consequently, SAR reception occurs across frequency bands allocated for other broadcasting stations and SAR often suffers radio frequency interference (RFI). This study proposes a novel approach to suppress RFI in SAR. The authors combine the classical spectral analysis method singular spectrum analysis, with an efficient subspace tracking method fast approximated power iteration (FAPI) which has the most economical computational cost among state-of-the-art algorithms. Furthermore, they supply a gap of FAPI so that it can achieve rank tracking based on eigenvalues distribution of the covariance matrix of the observed SAR record and in that way can detect the number of RFIs along the SAR azimuth time. The proposed algorithm is a time domain processing method and performs on SAR range records one-by-one. Finally, well-focused SAR imagery is obtained from the ‘decontaminated’ SAR data using classical SAR imaging algorithms. The proposed algorithm could be applied in complicated interference circumstances no matter the styles of RFI. Imaging results of artificial and real data have proved the validity and robustness of this algorithm.

Sparse microwave imaging: Principles and applications

This paper provides principles and applications of the sparse microwave imaging theory and technology. Synthetic aperture radar (SAR) is an important method of modern remote sensing. During decades microwave imaging technology has achieved remarkable progress in the system performance of microwave imaging technology, and at the same time encountered increasing complexity in system implementation. The sparse microwave imaging introduces the sparse signal processing theory to radar imaging to obtain new theory, new system and new methodology of microwave imaging. Based on classical SAR imaging model and fundamental theories of sparse signal processing, we can derive the model of sparse microwave imaging, which is a sparse measurement and recovery problem and can be solved with various algorithms. There exist several fundamental points that must be considered in the efforts of applying sparse signal processing to radar imaging, including sparse representation, measurement matrix construction, unambiguity reconstruction and performance evaluation. Based on these considerations, the sparse signal processing could be successfully applied to radar imaging, and achieve benefits in several aspects, including improvement of image quality, reduction of data amount for sparse scene and enhancement of system performance. The sparse signal processing has also been applied in several specific radar imaging applications.

SAR Processors Based on Signal or Interference Subspace Detector

In this paper, we present a new synthetic aperture radar (SAR) processor relying on subspace detectors. The scattering pattern of the constituent element of the man-made target (MMT) to detect is used to improve the detection capabilities. In the particular case of foliage penetration (FOPEN), we propose to reduce the probability of false alarms by considering the scattering pattern of almost vertical trunks (modeled here as dielectric cylinders). New subspace detectors are derived to look for either an interference or a target element in a noisy environment. A new processor based on these subspace detectors is successively applied to simulated data. In addition, results obtained with real data reveal a significant improvement in the detection capabilities both in comparison with a classical SAR processor and with the subspace signal detector SAR (SSDSAR) processor (without interference model).

4-D SAR Imaging: The Case Study of Rome

Four-dimensional synthetic aperture radar (SAR) imaging, also known as differential SAR tomography, is a new research topic in the framework of coherent multitemporal/multibaseline SAR processing that extends the interferometry concept. Four-dimensional SAR imaging-based processing could improve the capability of ground-scatterer monitoring with respect to classical differential interferometric SAR processing. The first results on the applicability of such an advanced tomographic SAR processing to real spaceborne data were recently discussed in the literature. In this letter, we present the results of an experiment with a data set that demonstrates the potentialities of this new technique for monitoring complex targets, such as infrastructures.

Fusion of Quickbird MS and RADARSAT SAR data for urban land-cover mapping: object-based and knowledge-based approach

The objective of this research is to evaluate Quickbird multi-spectral (MS) data, multi-temporal RADARSAT Fine-Beam C-HH synthetic aperture radar (SAR) data and fusion of Quickbird MS and RADARSAT SAR for urban land-use/land-cover mapping. One scene of Quickbird multi-spectral imagery was acquired on 18 July 2002 and five-date RADARSAT fine-beam SAR images were acquired during May to August 2002. Quickbird MS images and RADARSAT SAR data were classified using an object-based and rule-based approach. The results demonstrated that the object-based and knowledge-based approach was effective in extracting urban land-cover classes. For identifying 16 land-cover classes, object-based and rule-based classification of Quickbird MS data yielded an overall classification accuracy of 87.9% (kappa: 0.868). For identifying 11 land-cover classes, object-based and rule-based classification of RADARSAT SAR data yielded an overall accuracy: 86.6% (kappa: 0.852 ). Decision level fusion of Quickbird classification and RADARSAT SAR classification was able to take advantage of the best classifications of both optical and SAR data, thus significantly improving the classification accuracies of several land-cover classes (25% for pasture, 19% for soybeans, 17% for rapeseeds) even though the overall classification accuracy of 16 land-cover classes increased only slightly to 89.5% (kappa: 0.885).

Advanced fractal approach for unsupervised classification of SAR images

Unsupervised classification of Synthetic Aperture Radar (SAR) images is the alternative approach when no or minimum apriori information about the image is available. Therefore, an attempt has been made to develop an unsupervised classification scheme for SAR images based on textural information in present paper. For extraction of textural features two properties are used viz. fractal dimension D and Moran’s I. Using these indices an algorithm is proposed for contextual classification of SAR images. The novelty of the algorithm is that it implements the textural information available in SAR image with the help of two texture measures viz. D and I. For estimation of D, the Two Dimensional Variation Method (2DVM) has been revised and implemented whose performance is compared with another method, i.e., Triangular Prism Surface Area Method (TPSAM). It is also necessary to check the classification accuracy for various window sizes and optimize the window size for best classification. This exercise has been carried out to know the effect of window size on classification accuracy. The algorithm is applied on four SAR images of Hardwar region, India and classification accuracy has been computed. A comparison of the proposed algorithm using both fractal dimension estimation methods with the K-Means algorithm is discussed. The maximum overall classification accuracy with K-Means comes to be 53.26% whereas overall classification accuracy with proposed algorithm is 66.16% for TPSAM and 61.26% for 2DVM.

Multiscale modeling for classification of SAR imagery using hybrid EM algorithm and genetic algorithm

A novel method that hybridizes genetic algorithm (GA) and expectation maximization (EM) algorithm for the classification of synthetic aperture radar (SAR) imagery is proposed by the finite Gaussian mixtures model (GMM) and multiscale autoregressive (MAR) model. This algorithm is capable of improving the global optimality and consistency of the classification performance. The experiments on the SAR images show that the proposed algorithm outperforms the standard EM method significantly in classification accuracy.

Classification of crop and soil homogenous zones using multipolarization C-band SAR

With the acquisition of polarimetric radar data from satellites such as RADARSAT-2 and ALOS PALSAR, this research project explored the potential of using multipolarization C-band synthetic aperture radar (SAR) data to map homogenous soil and crop characteristics within agricultural fields. Homogenous zones within crop fields were defined using coincidental apparent soil electrical conductivity (EC) measured from a soil EC mapping system, green leaf area index (LAI) derived from airborne hyperspectral CASI data, and yield converted to phytomass at harvest (Ph) derived from a yield monitor. Airborne polarimetric C-band SAR images (5.6 cm) at 4 m spatial resolution were obtained over a test site in Ottawa, Canada, on 13 June, 26 June, and 19 July 2001. Linear (HH, VV, HV) and circular (RR, LL, RL) polarization images were synthesized from these data. Once the homogenous zones were determined by EC, LAI, and Ph, the multipolarization SAR data were classified based on these zones using single polarizations and combinations of linear and circular polarizations. Within-field classification accuracies of SAR data were higher when they were based on EC and LAI homogenous zones relative to Ph delineations. The lower SAR classification accuracies based on Ph zones were likely attributed to biomass zones defined relative to harvest rather than coincident with SAR data acquisition. LAI zones were better classified using early season SAR data when greater between-zone variability was present. EC homogenous zones were better classified using SAR data later in the growing season when soil variability was manifested in the vegetation response. Within-field classification performance over the season was dependent on the crop type and the incidence angle. At larger incidence angles (55°), results were more dependent on crop type when compared with smaller incidence angles (35°). In general, a single polarization was capable of discriminating many of the within-field homogenous zones, including EC and LAI zones in corn and LAI and Ph zones in wheat. The linear (HV) and circular (LL or RR) cross-polarizations were the best single polarizations to classify these homogenous zones. Within-field classification of the SAR into EC homogenous zones in the wheat field required polarization combinations, particularly combinations that included HH and VV polarizations. Circular polarizations, either as single polarizations or as part of a combination, were necessary to classify SAR data into EC or LAI homogenous zones as the season progressed. These results suggest that multipolarization SAR data are useful in classifying within-field homogenous zones of corn and wheat, but that this application will require access to data from sensors that synthesize circular polarizations.

Estimation of surface roughness and stone cover in burnt soils using SAR images

Abstract In many quartzitic slopes in the NW of the Iberian Peninsula there is a high topsoil stoniness which produces a random surface roughness. The surface roughness is one of the control parameters in the signal of Synthetic Aperture Radar (SAR) images. In this work we test the use of the SAR images for distinguishing and mapping areas with different surface roughness which, in this study, correspond with areas with different surface stoniness. In accordance with the geographic location, the relief of the study area and the aims of the project JERS and ENVISAT SAR images were used. The images are characterized by the high view angles, between 33° and 42°, alternating polarization (HH/HV, VV/VH) data and ascending and descending passes. In order to describe the soil roughness various measuring methods were applied in plot parcels (direct measurements of height of clasts and laser scanning) and different roughness models were tested. The results show that there are relation between SAR signal (DN) and the soil roughness classes: DN = aθb and the b parameter is correlated with the roughness. The best fit between roughness model and SAR data was found for mean standard deviation (rms) height of clasts. Once verified that the correlation between SAR signal and surface roughness exists, the ENVISAT images were classified into three successive surface roughness classes. In order to facilitate the classification Principal Components Analysis were carried out over all polarization bands and the object oriented classification was applied. To validate the accuracy of the SAR images classification we elaborated by photointerpretation a surface condition map for an area of 10,700 ha. The SAR classification and photointerpreted map were compared for three classes: surface without stones, surface with stones (low and medium stoniness) and total cover of loose clasts. The overall coincidence between the maps is about 70%. Besides, large areas located in the northwest of the Iberian Peninsula suffer soil degradation caused by wildfires and the increase of surface stoniness is one of the effects described in the postfire erosion. The influence of wildfires and the associated erosion is particularly important on quartzitic slopes partially covered by coarse grain-supported deposits, like the ones considered in this study. The available data for the area of study suggest the high correlation between the different classes of surface roughness and stoniness obtained from SAR imagery, and the gradual different status of postfire soil erosion.