Search

We propose a model for generating low-frequency synthetic aperture radar (SAR) clutter that relates model parameters to physical characteristics of the scene. The model includes both distributed scattering and large-amplitude discrete clutter responses. The model also incorporates the SAR imaging process, which introduces correlation among image pixels. The model may be used to generate synthetic clutter for a range of environmental operating conditions for use in target detection performance evaluation of the radar and automatic target detection/recognition algorithms. We derive a statistical representation of the proposed clutter model's pixel amplitudes and compare with measured data from the CARABAS-II SAR. Simulated clutter images capture the structure and amplitude responses seen in the measured data. A statistical analysis shows an order of magnitude improvement in model fit error compared with standard maximum-likelihood (ML) density fitting methods.

We consider characterization of building interior structure from two-pass interferometric circular synthetic aperture radar data. We model returns from complex building structures as the sum of the responses from scattering primitives--plates, dihedrals, and trihedrals--observed through transfer functions that characterize both the transmission through and reflection from dielectric surfaces. Maximum-likelihood estimates for the location of primitive features and wall propagation/ reflection parameters are obtained through a two-stage detection-and-estimation algorithm. We employ a sparse reconstruction algorithm to detect primitive signatures in the presence of wall effects and refine our estimates using a nonlinear minimization of local cost functions. We show that using simplified wall and primitive models, we can extract geometrically relevant features which can be combined to reveal 3-D floorplan estimates of buildings. Examples illustrate the effectiveness of the feature extraction approach. Index Terms--Buildings, feature extraction, maximum-likelihood (ML) estimation, radar imaging, radar signal processing.

The effect of motion measurement errors (MMEs) on measured bistatic synthetic aperture radar (SAR) phase history data that has been motion compensated to the scene origin is discussed. By deriving limits on the allowable MMEs to be used as system specifications, it is demonstrated that proper orientation of a bistatic SAR image during image formation process allows application of monostatic SAR autofocus algorithms in postprocessing to mitigate image defocus.

The design of planar arrays that optimize direction-of-approval (DOA) estimation performance is discussed. The optimal arrays have elements on the constraint boundary, thus providing a reduced dimension iterative solution procedure.

The polar format algorithm (PFA) for spotlight synthetic aperture radar (SAR) is based on a linear approximation for the differential range to a scatterer. We derive a second-order Taylor series approximation of the differential range. We provide a simple and concise derivation of both the far-field linear approximation of the differential range, which forms the basis of the PFA, and the corresponding approximation limits based on the second--order terms of the approximation.

Matched filtering (MF) of phase history data is a mathematically ideal but computationally expensive approach to bistatic synthetic aperture radar (SAR) image formation. Fast backprojection algorithms (BPAs) for image formation have recently been shown to give improved O([N.sup.2][log.sub.2]N) performance. In this paper, an O([N.sup.2][log.sub.2]N) bistatic polar format algorithm (PFA) based on a bistatic far-field assumption is derived. This algorithm is a generalization of the popular PFA for monostatic SAR image formation and is highly amenable to implementation with existing monostatic image formation processors. Limits on the size of an imaged scene, analogous to those in monostatic systems, are derived for the bistatic PFA.

Array geometries with direction-of-arrival (DOA) estimation performance which are isotropic are investigated. An isotropic array is one whose Cramer-Rao bound (CRB) on the DOA of a single source is uniform for all angles.

An algorithm is presented for the simultaneous estimation of parameters for a noisy exponential noise sequence.

The focus of this paper is a convergence study of the frequency sensitive competitive learning (FSCL) algorithm. We approximate the final phase of FSCL learning by a diffusion process described by a Fokker-Plank equation. Sufficient and necessary conditions are presented for the convergence of the diffusion process to a local equilibrium. The analysis parallels that by Ritter and Schulten for Kohonen's self-organizing map (SOM). We show that the convergence conditions involve only the learning rate and that they are the same as the conditions for weak convergence described previously. Our analysis thus broadens the class of algorithms that have been shown to have these types of convergence characteristics. Index Terms - Diffusion processes, Fokker-Plank equations, neural networks, learning systems, vector quantization.

A large-sample maximum likelihood (ML) algorithm for estimating the directions of arrival (DOA's) and signal amplitudes of known, possibly coherent signals impinging on an array of sensors is shown. The algorithm is an extension of the DEML technique of Li et al. that handles the coherent multipath that may be exist in the signals. The algorithm is computationally efficient because the nonlinear minimization step decouples into a set of minimizations of smaller dimension.

Physics-based parametric radar scattering models help in estimating attributes of scattering centers, which in turn help in analyzing synthetic aperture radar data for automatic target recognition. The geometrical theory of diffraction is used to predict the scattering parameters. Scattering geometry, polarization response, aspect-dependent amplitude taper and location are the main estimated attributes. A least square algorithm for feature estimation and a statistical analysis of the scattering model are described. Two reduced complexity scattering models are discussed.

A maximum likelihood (ML) approach is used in the array signal processing for stochastic coherent signal sources. For large samples, the ML estimation of signal parameters does not require the knowledge of signal covariance matrix ranks. The asymptotic nature of ML estimates of the noise power and signal covariance matrix is analyzed.

Incorporation of data decimation in TLS-Prony method decreases computational complexity of the method by introducing several low-order estimations. Statistical analysis of pole variance of the modified TLS-Prony method helps compare parameter estimation accuracy as a function of decimation factors. Use of decimation procedure produces similar statistical performance results at low computation cost compared to traditional TLS-Prony algorithms.

A new method using one-dimensional (1-D) total least squares facilitates the determining of amplitude coefficients and exponential modes in a two-dimensional (2-D) Prony model. The method consists of two parts which use 1-D singular value decomposition-based methods for parameter estimation. The frequencies at any point within the 2-D frequency plane are located by the algorithms involved in this new method.

Complete formulations for the Cramer-Rao bound of the parameters of an exponential model with one set of poles and multiple sets of amplitude coefficients are presented. Results show that the polar and relative polar Cramer-Rao bounds are unaffected by the absolute phase of the poles and amplitude coefficients. Meanwhile, in rectangular parameters, the real and imaginary part bounds are equal and uncorrelated.

The performance of the modified sample matrix inverse algorithm is evaluated based on the K-snapshots estimates of thediagonally loaded covariance matrix. O(1/K) approximations of the statistics ofthe array weights, output powers and output power ratios are derived as a function of the amount diagonal loading and the number of K-snapshots used in the covariance matrix estimate. The results of the study are validated by MonteCarlo simulations.

A study on the bispectral processing of complex-valued signals is presented. Formulas for thedifferent signal bispectra that may be obtained through a variation of the complex conjugate's position in the third-order cumulant are also given, together with their corresponding correlations. A study on the use of third-order cumulants for parametric modeling of such signals shows the need for a different method as compared to real-valued signal modeling.

A high resolution modeling technique using the Prony algorithm was developed for radar target description. The Prony model, a signal processing method, was improved by the application of singular value decomposition. It then estimated radar target scattering centers from frequency domain data taken at compact range at the ElectroScience Laboratory of the Ohio State University. Inverse scattering is an approach to map the profile of an object through its approximation as a set of discrete scattering centers. Test showed that a change of 15 degrees by the target led to predictable estimation of data response.

A Bayesian approach is presented for model-based classification of images with application to synthetic-aperture radar. Posterior probabilities are computed for candidate hypotheses using physical features estimated from sensor data along with features predicted from these hypotheses. The likelihood scoring allows propagation of uncertainty arising in both the sensor data and object models. The Bayesian classification, including the determination of a correspondence between unordered random features, is shown to be tractable, yielding a classification algorithm, a method for estimating error rates, and a tool for evaluating performance sensitivity. The radar image features used for classification are point locations with an associated vector of physical attributes; the attributed features are adopted from a parametric model of high-frequency radar scattering. With the emergence of wideband sensor technology, these physical features expand interpretation of radar imagery to access the frequency- and aspect-dependent scattering information carried in the image phase. Index Terms--Model-based classification, parametric modeling, point correspondence, radar image analysis.

Within the hyperspectral community, change detection is a continued area of interest. Interesting changes in imagery typically correspond to changes in material reflectance associated with pixels in the scene. Using a physical model describing the sensor-reaching radiance, change detection can be formulated as a statistical hypothesis test. Complicating the problem of change detection is the presence of shadow, illumination, and atmospheric differences, as well as misregistration and parallax error, which often produce the appearance of change. The proposed physical model incorporates terms to account for both direct and diffuse shadow fractions to help mitigate false alarms associated with shadow differences between scenes. The resulting generalized likelihood ratio test (GLRT) provides an indicator of change at each pixel. The maximum likelihood estimates of the physical model parameters used for the GLRT are obtained from the entire joint data set to take advantage of coupled information existing between pixel measurements. Simulation results using synthetic and real imagery demonstrate the efficacy of the proposed approach. [ABSTRACT FROM AUTHOR]

This paper discusses reconstruction of three-dimensional surfaces from multiple bistatic synthetic aperture radar (SAR) images. Techniques for surface reconstruction from multiple monostatic SAR images already exist, including interferometric processing and stereo SAR. We generalize these methods to obtain algorithms for bistatic interferometric SAR and bistatic stereo SAR. We also propose a framework for predicting the performance of our multistatic stereo SAR algorithm, and, from this framework, we suggest a metric for use in planning strategic deployment of multistatic assets. [ABSTRACT FROM AUTHOR]

Details a study which presented a Bayesian approach for model-based classification of images with application to synthetic-aperture radar. Physical model for sensor data and feature extraction; Hypothesis testing; Example of Bayes classification.

Presents a convergence study of the frequency sensitive competitive learning (FSCL) algorithm. Approximation of the final phase of FSCL learning; Demonstration of how the convergence conditions involve only the learning rate; Comment that the analysis expands the class of algorithms that have been shown to have convergence characteristics.