Search

The paper presents two algorithms for texture primitive feature extraction on Single Look Complex (SLC) and Polarimetric Synthetic Aperture Radar (PolSAR) SLC data. We assume the data to be modeled by a Gauss-Markov Random Field (GMRF): a complex GMRF model for characterizing the spatial correlation in SLC data and an extension of the model for inter-band correlation characterization. The complex GMRF characterizes the spatial relationship of a two-dimensional complex signal, i.e. SLC SAR data. The extended model characterizes the spatial interaction and the inter-band pixels correlation between the polarimetric complex channels. The Bayesian approach permits to deal with model fitting and selection in a direct way. The results are presented on a polarimetric E-SAR L band scene of Mannheim, Germany.

This paper deals with the registration of near-real-time satellite SAR images by an automatic image-to-image matching approach. The approach is dedicated to crisis mapping applications where the task is to register near real-time (NRT) SAR images within some hours against a reference image. Our registration strategy consists of the following steps: Feature extraction and matching between a NRT image and an image with precise orbit. The resultant points of the reference image are used as ground control points (GCPs) to adjust the SAR imaging geometry. Then, geocoding of the NRT SAR image can be carried out. The main task is to find a feature extractor which performs well on SAR data. We use a new point operator for feature extraction to extract scale- and rotation-invariant features (SIFT features). The results show that the approach can deal with a wide range of scenes even with different SAR sensors and different incidence angles.

Segmentation algorithms are often used in many image processing applications like compression, restoration, content extraction, and classification. In particular as for content extraction works carried out in the past decade have demonstrated that multi-frequency fully polarimetric SAR observations are particularly interesting, thanks to physical properties of the backscattered signal at various frequencies and polarizations. To achieve a good classification, the main difficulty is that SAR images are often embedded in heavy speckle. Segmentation of multi/hyperspectral (optical) imagery is obtained by means of algorithms based on image models, which exploit the spatial dependencies of land-covers. Unfortunately, speckle noise hides such spatial dependencies in observed SAR data. With the aim of investigating on a content extraction algorithm capable of discriminating cover classes present in the observed SAR image, heterogeneity features are used here to emphasize spatial dependencies in the data. Thus, observed pixel values are mapped into features, that take "similar" values on "similar" textures. This allows for using the same procedure of the optical case. Obviously, homogeneity/heterogeneity feature and segmentation quality are fundamental for classification accuracy. Here, the problem is tackled through the joint use of information theoretic SAR features and of a segmentation algorithm based on Markov Random Fields (MRFs).

Multi-temporal analysis of Synthetic Aperture Radar (SAR) images has gained an ever increasing attention due to the availability of several satellite platforms with different revisit times and to the intrinsic capability of the SAR system of producing all-weather observations. As a drawback, automated analysis in general and change detection in particular are made difficult by the inherent noisiness of SAR imagery. Even if a pre-processing step aimed at speckle reduction is adopted, most of algorithms borrowed from computer vision cannot be profitably used. In this work, a novel pixel feature suitable for change analysis is derived from information-theoretic concepts. It does not require preliminary de-speckling and capable of providing accurate change maps from a couple of SAR images. The rationale is that the negative of logarithm of the probability of an amplitude level in one image conditional to the level of the same pixel in the other image conveys an information on the amount of change occurred between the two passes. Experimental results carried out on two couples of multi-temporal SAR images demonstrate that the proposed IT feature outperform the Log-Ratio in terms of capability of discriminating changes.

Spatial characterization is a recurrent problem in flood risk management. However, spatial characteristics of a flood event are not easily gaugeable directly in the field with current tools. Techniques based on remote sensing imagery, especially on satellite imagery, could be very useful to complete gauging, calculate flood extends and volumes. RADAR sensors, with their exclusive cloud penetration capacity, appear to be of great interest in flood events monitoring. The general aim of our work is to develop way for spatial characterisation of flood events, using RADAR images, in order to help hydraulic modelling. Therefore, we have chosen a three-stage approach. The first stage aims to extract flooded area from RADAR images, the second one to enhance quality of inundated area calculation and to determine waterline elevation, and the last one to use spatial information calculated for flood modelling. Here, we just present the result for the two first stages. RADAR data, which are imperfect and sometimes incomplete, allow us to product only uncertain flood maps. As a consequence, to relevantly extract flooded area using RADAR images, we have to focus on accuracies and uncertainties due to RADAR data and treatments used for mapping, we have to examine uncertainties linked to RADAR data and. These uncertainties are especially linked with RADAR characteristics (e.g.: speckle, spatial and radiometric resolutions, etc.), meteorological conditions (e.g.: wind, rain) and land cover (e.g.: forest and cities). This stage especially enables us to share calculated flood extent into relevant and irrelevant areas. To enhance the calculated flood extent and determine water levels, we match relevant flooded areas with very high resolution DEM thanks to a Geographic Information System (GIS). Moreover, to reduce water level uncertainties and to make water level estimations hydraulically coherent we introduce rules of hydraulic coherence on successive water levels along flood plain. The perspective of these to first steps is to use spatial characteristics calculated with SAR images and DEM to help hydraulic modelling. For example, we aim to use water level estimations as initial condition, or for calibration of a hydraulic model. This kind of improvement of SAR imagery can be very useful to develop automatic methods, which could allow us to calculate future evolution of a beginning flood. The area of interest is the Moselle river between Thionville and Berg-sur-Moselle (France). The accuracy of water level estimations is about ± 50 cm using RADARSAT 1 image of flood (acquired 02-28-1997).

In this paper, we propose a new mathematical model for detecting in an image singularities of codimension greater than or equal to two. This means we want to detect points in a 2-D image or points and curves in a 3-D image. We drew one's inspiration from Ginzburg-Landau (G-L) models which have proved their efficiency for modeling many phenomena in physics. We introduce the model, state its mathematical properties and give some experimental results demonstrating its capability.

Synthetic aperture radar (SAR) imagery has found important applications due to its clear advantages over optical satellite imagery one of them being able to operate in various weather conditions. However, due to the physics of the radar imaging process, SAR images contain unwanted artifacts in the form of a granular look which is called speckle. The assumptions of the classical SAR image generation model lead to a Rayleigh distribution model for the histogram of the SAR image. However, some experimental data such as images of urban areas show impulsive characteristics that correspond to underlying heavy-tailed distributions, which are clearly non-Rayleigh. Some alternative distributions have been suggested such as the Weibull, log-normal, and the k-distribution which had success in varying degrees depending on the application. Recently, an alternative model namely the /spl alpha/-stable distribution has been suggested for modeling radar clutter. In this paper, we show that the amplitude distribution of the complex wave, the real and the imaginery components of which are assumed to be distributed by the /spl alpha/-stable distribution, is a generalization of the Rayleigh distribution. We demonstrate that the amplitude distribution is a mixture of Rayleighs as is the k-distribution in accordance with earlier work on modeling SAR images which showed that almost all successful SAR image models could be expressed as mixtures of Rayleighs. We also present parameter estimation techniques based on negative order moments for the new model. Finally, we test the performance of the model on urban images and compare with other models such as Rayleigh, Weibull, and the k-distribution.

In this paper, LLMMSE filtering is performed in the undecimated wavelet domain by means of an adaptive rescaling of detail coefficients. The amplitude of each coefficient is divided by the variance ratio of the noisy coefficient to the noise-free one. All the above quantities are analytically calculated from the speckled image, the speckle variance, and the wavelet filters only, without assuming any model to describe the underlying backscatter. Empirical criteria based on distributions of multiresolution C/sub v/ calculated in the undecimated wavelet domain are introduced to mitigate the rescaling of coefficients in highly heterogeneous areas where the speckle is not fully developed, to definitely avoid the slight blurring noticed in some textures and the perceivable smearing of point targets. Experiments carried out on SAR images demonstrate that the visual quality of results is excellent in terms of both background smoothing and preservation of edge sharpness, textures, and point targets. The absence of decimation in the wavelet decomposition avoids the typical impairments produced by critically-subsampled wavelet-based denoising.

A suited approach to calculate the fractal dimension of images comes from power spectra fractional Brownian motions: the ratio between powers at different scales is related to the persistence parameter H and, thus, to the fractal dimension D = 3 - H. The signal-dependent nature of speckle noise, however, prevents from a correct estimation of fractal dimension from Synthetic Aperture Radar (SAR) images. Here, we propose and assess a novel method to obtain D based on the multiscale decomposition provided by the normalized Laplacian pyramid (NLP), which is a bandpass representation obtained by dividing the layers of an LP by its expanded baseband and designed to yield noise that is signal-independent. Experiments on both synthetic and true SAR images corroborate the underlying assumptions.

A filter aimed at denoising should strongly smooth uniform regions, while preserving edges. On textured areas, the filter should attain a compromise to achieve some enhancement without destroying useful information. Filtering performances, however, locally depend on the statistical characteristics of the imaged signal, which can be embodied by the concept of local heterogeneity. It is shown that statistically homogeneous regions originate clusters in the scatterplot of standard deviation to mean. Textured regions yield scatterpoints spread apart to a larger extent. Edges produce outliers. Thus, the homogeneity may be locally measured from the joint PDF of estimated local standard deviation to estimated local mean. For each pixel having a measured local mean and a measured standard deviation, a point is detected in the PDF plane: the corresponding density is taken as a measure of homogeneity of that pixel. In this work a hybrid filter, i.e., a set of filters is considered, the denoising capability of each of which depends on the degree of local homogeneity. Images are individually processed by each filter, and the filtered image is obtained by switching among such channels at each pixel position, based on thresholding a heterogeneity feature in order to identify a number of classes, for each of which the noise-free image signal is best estimated by one of the filters. Visual judgments on simulated noisy images agree with this tendency.

Speckle reduction in synthetic aperture radar (SAR) images is a key point to facilitate applicative tasks. A filter aimed at de-speckling should energetically smooth homogeneous regions, while preserving point targets, edges, and linear features. A trade-off, however, should be arranged on textured areas. Filtering capabilities depend on local image characteristics, and generally no filter outperforms the others in every situation. In this work, a set of adaptive filters is considered with attention to those oriented towards a multiresolution approach. Images are individually processed by each filter, and the output at each pixel position is obtained by choosing one out of the channels. The selection is based on thresholding a pixel feature based on a novel definition of inhomogeneity for SAR images and capable of describing both textured areas and ideal scatterers. Results on true SAR images show that also an empirical choice of thresholds is non-critical: the proposed scheme outperforms each filter individually, at least according to visual criteria.

We propose an original approach to near-lossless compression, based on quantizing and encoding the Rational Laplacian Pyramid (RLP) of a SAR image. The RLP features the desirable property that the noise on its layers, except for the base-band, is practically independent of the signal. The RLP has mean roughly equal to one and variance proportional to that of the speckle, sigma(u)**2, on homogeneous areas, and larger elsewhere. The base-band icon is encoded, and the bottom layer of the RLP, diminished by its expected value, is uniformly quantized with a step size Delta = sigma(u). The steps on the other layers are arbitrary because of the quantization noise feedback loop at the encoder. Whenever larger compression ratios are desired, instead of increasing the Delta, a dead-zone can be introduced in the quantizer: pyramid coefficients lying below the threshold (half of the dead-zone) are set at zero; those above are quantized and encoded. The procedure is equivalent to a hard thresholding, a widely used technique for image de-noising in the wavelet domain. Experimental results show that extremely high-quality images are obtained at reasonable rates; as the compression is increased, images are progressively de-noised to be accommodated in the available rate.

A well-suited approach to calculate the fractal dimension of digital images stems from the power spectrum of a fractional Brownian motion: the ratio between powers at different scales is related to the persistence parameter H and, thus, to the fractal dimension D = 3 - H. The signal-dependent nature of the speckle noise, however, prevents from a correct estimation of fractal dimension from Synthetic Aperture Radar (SAR) images. Here, we propose and assess a novel method to obtain D based on the multiscale decomposition provided by the normalized Laplacian pyramid (NLP), which is a bandpass representation obtained by dividing the layers of an LP by its expanded baseband and is designed to force the noise to become signal-independent. Extensive experiments on synthetic fractal textures, both noise-free and noisy, corroborate the underlying assumptions and show the performances, in terms of both accuracy and confidence of estimation, of pyramid methods compared with the well-established method based on the wavelet transform. Preliminary results on true SAR images from ERS-1 look promising as well.

An unsupervised method is first proposed to assess the variance and the spatial correlation coefficients of speckle noise in SAR images. They are obtained as regression coefficients, the former of local standard deviation to local mean, the latter of local unity-lag covariance to local variance; both calculated on homogeneous areas. For this purpose, an automatic procedure has been developed, based on that homogeneous areas produce clusters of scatterpoints that are aligned along the regression line. On true SAR images, the method is capable to carefully reject textured regions, in which speckle may be not fully developed and the variance of the signal is not negligible. On simulated speckled images, an impressive accuracy is obtained. Once the noise parameters are known, adaptive filtering is applied in a multiresolution fashion, to take advantage of increasing SNR of the noisy image at increasing scales, as well as to cope with the spatial correlation of the noise that is halved together with the resolution. Laplacian pyramids are generalized to the noise model by defining ratios of combinations of lowpass image versions, in which the dependence of the noise on the signal is largely removed, together with the nonstationarity of the mean. Experiments on both real and synthetic images demonstrate a high accuracy of results, both for noise estimation and for filtering.

It is widely recognized that SAR images exhibit a fractal behavior represented by the concept of fractal dimension, which is related to an intuitive concept of surface "roughness". The most suited approach to compute the fractal dimension comes from the power spectra of a fractal Brownian motion: the ratio between energies at different scales is related to the persistence parameter H and, thus, to the fractal dimension D equals 3 - H. The signal-dependent nature of speckle, however, prevents from the exploitation of this property to estimate the fractal dimension of SAR images. In this paper, we propose and assess a novel method to obtain such a fractal signature, based on the multi-scale image decomposition provided by the normalized Laplacian pyramid (NLP), which is a bandpass representation obtained by dividing the layers of the LP by expanded versions of its baseband, designed to exhibit noise that is independent of the signal. Thus, by analyzing SAR image texture on multiple scale through the NLP, it is possible to highlight and assess fractal behaviors of the radar cross-section. Experiments on both synthetic and true SAR images corroborate the theoretical assumptions underlying the proposed approach.

Local-statistics speckle filtering has been extended to multitemporal SAR data by exploiting the temporal correlation of the speckle noise across a set of images of the same scene taken at different times. A recursive nonlinear transformation aimed at decorrelating the data across time, while retaining the multiplicative noise model, is defined from the geometric means and the ratios of couples of spatially overlapped observations. The temporal correlation coefficient (TCC) is estimated from the modes of the distributions of the local variation coefficient Cv computed on transformed couples of images. The images are filtered in the transformed domain and reversely transformed to yield despeckled observations in which seasonal changes are preserved, or even highlighted, and texture analysis is expedited. Tests on four SAR images from repeat-pass ERS-1 corroborate the theoretical assumptions and show the filtering performances of the proposed approach.

Speckle filtering in multiresolution frameworks is the key to extend spatial adaptivity also across scales. Several procedures based on wavelets and pyramids, in which the different resolutions are adaptively filtered, have been introduced for selective speckle smoothing with preservation of features and point targets. In this work, comparison tests on true and synthetic speckled images show significant advances of the multi- over the single-scale approach, especially for the pyramid based scheme proposed by the authors, where visual comparisons on SAR images show better texture preservation.

Speckle reduction in synthetic aperture radar (SAR) images is a key point to facilitate applicative tasks. A filter aimed at speckle reduction should energetically smooth homogeneous regions, while preserving point targets, edges, and linear features. A tradeoff, however, should be arranged on textured areas. Filtering capabilities depend on local image characteristics, and generally no filter outperforms the others in every situation. In this work, a set of adaptive filters is considered with attention to those oriented towards a multiresolution approach. Images are individually processed by each filter, and the output at each pixel position is obtained by choosing one out of the channels. The selection is based on thresholding local features accounting for both space-varying statistics and geometry. Results on true SAR images show that also an empirical choice of thresholds is noncritical: the novel scheme outperforms each filter individually, at least according to visual criteria.

A multiresolution filtering procedure is derived from the framework of Laplacian pyramids and proposed for edge-preserving MMSE smoothing of images affected by multiplicative noise with speckle statistics. The basic idea of the novel scheme is to apply an MMSE filter (e.g., Kuan's filter) to the different resolution layers in which the noisy image is decomposed. A noise-free image version is achieved by recombining the layers of the pyramid, once they have been filtered. Comparison tests performed on images affected by synthetic speckle, with both one-look and multi-look statistics, show significant SNR improvements over the basic Kuan's scheme, resulting in far lower distortion between noise-free and processed images. Visual comparisons on true and simulated SAR images corroborate the objective results.

A recently investigated approach to noise filtering in digital images consists of considering a multiresolution decomposition of the input image, and applying a different adaptive filter to each resolution layer. The wavelet decomposition has been employed for multiresolution noise- reduction, thanks to its capability to capture spatial features within frequency subbands. Conversely, Laplacian pyramids (LP) look attractive because of their full band- pass frequency property, which enables connected image structures to be represented on multiple scales. The idea of the present work is to apply an adaptive minimum mean squared error filter to the connectivity-preserving different resolution layers into which the noisy image is decomposed. For natural images, each layer of the LP is characterized by a signal-to-noise ratio (SNR) that decreases for increasing spatial resolution. Therefore, each filter may be tuned to the SNR of the related layer, so as to preserve the spatial details of the less noisy layers to a larger extent. Once all the resolutions, including the base-band, have been adaptively smoothed, a noise-filtered image version is achieved by recombining the layers of the LP. Theoretical frameworks are developed for both additive and multiplicative noise models. Experimental results of de- noising carried out on images with simulated noise and on true synthetic aperture radar images validate the potentiality of the approach in terms of both SNR improvement and visual quality.

Spaceborne/airborne synthetic aperture radar (SAR) systems provide high resolution two-dimensional terrain imagery. The paper proposes a technique for combining multiple SAR images, acquired on flight paths slightly separated in the elevation direction, to generate high resolution three-dimensional imagery. The technique could be viewed as an extension to interferometric SAR (InSAR) in that it generates topographic imagery with an additional dimension of resolution. The 3-D multi-pass SAR imaging system is typically characterised by a relatively short ambiguity length in the elevation direction. To minimise the associated ambiguities we exploit the relative phase information within the set of images to track the terrain landscape. The SAR images are then coherently combined, via a nonuniform DFT, over a narrow (in elevation) volume centred on the 'dominant' terrain ground plane. The paper includes a detailed description of the technique, background theory, including achievable resolution, and the results of an experimental study.

Land use/cover change detection is very important in the application of remote sensing. In the case of Synthetic Aperture Radar (SAR) acquisitions for change detection, the standard detector or change measure is based on the ratio of images. However, this measure is sensitive to the speckle effect. In this paper, we improve change detection methods using a new change measure. The measure uses a grey level gradient or intensity information and the fractal dimension. The proposed measure is partitioned into two distinct regions, namely, changed and unchanged, using some change detection methods like Support Vector Machines (SVM), Fuzzy C -Means clustering (FCM) and artificial neural networks with a back propagation training algorithm. Experiments over the study area show that the results of implementing change detection methods are improved by using the proposed measure, in comparison to the classical log-ratio image. Also, results prove that the measure is very robust to the speckle effect.

A novel method is proposed to automatically measure the temporal correlation coefficient (TCC) of speckle from a set of SAR images of the same scene taken at different times. The knowledge of the TCC of speckle may expedite the detection and assessment of seasonal changes having occurred. In addition, it provides an upper bound to the maximum SNR increase achievable by multitemporal processing. A nonlinear transformation aimed at decorrelating the data across time, while retaining the multiplicative noise model, is defined. The TCC is estimated from the modes of the distributions of the local variation coefficient (C/sub v/) computed on the transformed images. Tests on SAR images (both synthetic and from ERS-1) show the accuracy and the robustness of the method, whose results are unaffected by underlying seasonal changes occurring across observations.

We present an approach for spatio-temporal co-registration of dynamic objects in Synthetic Aperture Radar (SAR) and optical imagery. Goal of this work is the performance evaluation of vehicle detection and velocity estimation from SAR images when comparing it with reference data derived from aerial image sequences. The results of evaluation show the challenges of traffic monitoring with SAR in terms of detection rates for individual vehicles.

Adding invariance properties to a dictionary-based model is a convenient way to reach a high representation capacity while maintaining a compact structure. Compact dictionaries of patches are desirable because they ease semantic interpretation of their elements (atoms) and offer robust decompositions even under strong speckle fluctuations. This paper describes how patches of a dictionary can be matched to a speckled image by accounting for unknown shifts and affine radiometric changes. This procedure is used to build dictionaries of patches specific to SAR images. The dictionaries can then be used for denoising or classification purposes.