Shawn M. Verbout

Template-based SAR ATR performance using different image enhancement techniques

The Lincoln Laboratory baseline ATR system for synthetic aperture radar (SAR) data applies a super-resolution technique known as high-definition vector imaging (HDVI) before the input image is passed through the final target classification subsystem. In previous studies, it has been demonstrated that HDVI improves target recognition performance significantly. Recently, however, several other viable SAR image enhancement techniques have been proposed and discussed in the literature which could be used in place of (or perhaps in conjunction with) the HDVI technique. This paper compares the performance achieved by the Lincoln Laboratory template-based classification subsystem when these alternative image enhancement techniques are used instead of the HDVI technique. In addition, empirical evidence is presented suggesting that target recognition performance could be further improved by fusing the classifier outputs generated by the best image enhancement techniques.

Parameter estimation for autoregressive Gaussian-mixture processes: the EMAX algorithm

The problem of estimating parameters of discrete-time non-Gaussian autoregressive (AR) processes is addressed. The subclass of such processes considered is restricted to those whose driving noise samples are statistically independent and identically distributed according to a Gaussian-mixture probability density function (pdf). Because the likelihood function for this problem is typically unbounded in the vicinity of undesirable, degenerate parameter estimates, the maximum likelihood approach is not fruitful. Hence, an alternative approach is taken whereby a finite local maximum of the likelihood surface is sought. This approach, which is termed the quasimaximum likelihood (QML) approach, is used to obtain estimates of the AR parameters as well as the means, variances, and weighting coefficients that define the Gaussian-mixture pdf. A technique for generating solutions to the QML problem is derived using a generalized version of the expectation-maximization principle. This technique, which is referred to as the EMAX algorithm, is applied in four illustrative examples; its performance is compared directly with that of previously proposed algorithms based on the same data model and that of conventional least-squares techniques.

A separation theorem for periodic sharing information patterns in decentralized control

Optimal decentralized control of a discrete-time stochastic system is considered under a periodic sharing information pattern. In this scenario, controllers share information with one-step delay every K time steps. The periodic sharing pattern is a generalization of the previously studied one-step delay sharing pattern, which is known to possess a nonclassical separation property. It is proven that the periodic sharing pattern has an analogous separation property.

New image features for discriminating targets from clutter

In this paper, we introduce a new set of image features for use in the discrimination algorithm of a baseline automatic target recognition (ATR) system. These new features are designed to capture the changes in spatial dispersion of the high-intensity pixels in the input image as the image is threshold at different intensity levels. We show that significantly better performance can be obtained when the new features are used in place of the baseline discrimination features. In particular, we demonstrate with a large set of high-resolution synthetic aperture radar imagery that, when the probability of detection is between 0.5 and 1.0, the false alarm density obtained using the new features is approximately 30 to 50 times lower than that obtained using the baseline features. For medium-resolution imagery, the false alarm density has been reduced by a factor of 3 to 5 using the new features.

Polarimetric techniques for enhancing SAR imagery

This paper describes several techniques for enhancing fully polarimetric, high-resolution synthetic aperture radar (SAR) imagery. Image enhancement is viewed as a pre-processing step to prepare the SAR data for sophisticated detection, discrimination, and classification algorithms. The paper considers enhancement techniques that use polarimetric information in the imagery to achieve one or both of the following goals: (1) reduction of image speckle, and (2) improvement in target-to-clutter contrast. Three enhancement techniques are presented: the polarimetric whitening filter (PWF), the polarimetric enhancement filter (PEF), and the polarimetric matched filter (PMF). These polarimetric processing techniques are applied to actual SAR data gathered by the Lincoln Laboratory MMW airborne sensor. The resulting target and clutter statistics are compared for typical data sets at both 1ft by 1ft resolution and 1m by 1m resolution.

High resolution radar target identification

The application of neural networks to the synthetic aperture radar (SAR) automatic target recognition (ATR) problem is discussed. In particular, initial studies investigating the use of an ART-2 self organizing neural-network in a 2-D SAR ATR system are summarized. The performance of the new neural net pattern-matching algorithm is compared with that of the baseline correlation pattern matching algorithm developed previously. This comparison includes evaluating the ability of the pattern matcher to reject nontargets (clutter discretes) and to classify the remaining detections into tank/APC/Howitzer categories.<<ETX>>

Two- and three-dimensional radar imaging at close range to a synthetic aperture

Summary form only given. Lincoln Laboratory is developing an ultra-wideband imaging radar that will be capable of two- and three-dimensional imaging at very close range to a synthetic aperture. The radar is fully coherent over two bandwidths (0.1 to 2 GHz and 2 to 18 GHz) and will be used for target imaging at X and Ku band, and for foliage penetration measurements over the VHF, UHF, and L bands. The radar is a portable scatterometer based on a Hewlett-Packard HP 8510C network analyzer combined with an HP 8360 frequency synthesizer and a Digital MicroVAX III computer. The characteristics of the radar are described, and technical issues arising in the analysis of the close-range imaging techniques are discussed.<<ETX>>

Multidimensional radar imaging with a polarimetric focused-beam scatterometer

A small millimeter-wave, polarimetric, focused-beam radar scatterometer known as the flashlight radar (FLR) has been developed. The FLR has three-dimensional resolution capability. It obtains high angle-angle resolution from a focused horn-lens antenna that produces a beam spot size of 4.6 in. in diameter at a distance of 10 ft from the radar. Six-inch range resolution is achieved from its 1-GHz swept bandwidth. The scatterometer operates at a center frequency of 35 GHz and uses the FM-CW method. It has been used to look at small targets and materials, and to make three-dimensional radar images of large objects. The radar output can be processed to simulate larger beamwidths, for example, or to give differing resolutions, or one of its three dimensions can be collapsed so as to compare an existing synthetic aperture radar image with that obtained from the scatterometer. The authors outline the FLR design, calibration, and processing. Also presented are samples of processed radar data obtained with the instrument.<<ETX>>

Backscatter theory and measurements for a millimeter-wave focused-beam radar

A portable millimeter-wave, fully polarimetric, high-resolution radar that uses a focused scalar horn-lens antenna has been developed. The one-way, copolarized beam in the focal plane has a measured -3 dB spot size diameter (circularly symmetric) of 4.6 in. The focused-beam properties of the antenna give the radar high angle-angle (azimuth and elevation) resolution, and for this reason the radar has been named the flashlight radar (FLR). The authors discuss the properties of the FLR as a scatterometer. The ability of the FLR to measure radar cross sections, taking into account the effects of the focused-beam antenna, is determined by modeling and measurements. A family of calibration curves is obtained and will be incorporated into the radar signal processing software. Backscatter measurements of two disks with radii of 2.8 and 5.0 cm were made with FLR at ranges from 1 to 6 m. The measurements agree very favorably with the theoretical predictions.<<ETX>>