Frederick D. Garber

SVM classifier applied to the MSTAR public data set

Support vector machines (SVM) are one of the most recent tools to be developed from research in statistical learning theory. The foundations of SVM were developed by Vapnik, and are gaining popularity within the learning theory community due to many attractive features and excellent demonstrated performance. However, SVM have not yet gained popularity within the synthetic aperture radar (SAR) automatic target recognition (ATR) community. The purpose of this paper is to introduce the concepts of SVM and to benchmark its performance on the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set.

Identifying moving HRR signatures with an ATR belief data association filter

The goal of this paper is to demonstrate the benefits of a tracking and identification algorithm that uses a belief data association filter for target recognition. By associating track and ID information, the belief filter accumulates evidence for classifying High-Range Resolution (HRR) radar signatures from a moving target. A track history can be utilized to reduce the search space of targets for a given pose range. The technique follows the work of Mitchell and Westerkamp by processing HRR amplitude and location feature sets. The new aspect of the work is the identification of multiple moving targets of the same type. The conclusions from the work is that moving ATR from HRR signatures necessitates a track history for robust target ID.

Characterization of ATR performance evaluation

This paper is concerned with the development of a framework in which reasonable bounds and approximations for the performance of automatic target recognition (ATR) systems may be obtained. The relative merits of evaluations focusing on various components of ATR systems are discussed. Techniques in which structural knowledge of the radar target (embodied in phenomenological and pattern models) might be used to characterize representations in pattern space, and distances in decision space are described. Preliminary results corresponding to various assumptions regarding structural constraints are presented.

Evaluation of nonparametric discriminant analysis techniques for radar signal feature selection and extraction

The applicability of discriminant-function-based feature selection techniques to the design of radar target classification systems is addressed. The characteristics and limitations of several nonparametric feature selection criteria based on discriminant functions are discussed from the point of view of developing systems for classifying targets based on their real-aperture radar returns. The results of applying feature selection criteria to a radar target database are presented. The classification performance of the selected feature sets is evaluated by means of Monte Carlo simulation studies.

Simple estimates of ATR performance and initial comparisons for a small data set

In this paper,we present an initial development of simple approximations for the performance of real-aperture radar ATR systems. In particular, we develop estimates of target separability using channel-capacity-like approximations, based on sensor constraints, and combinatoric-driven approximations, based on constraints imposed by the target. We then use these approximations to form estimates of the classification error probability for the ATR system. Finally, these estimates are compared with classification results for a small data set containing both synthetic and measured down-range responses form aircraft objects.

Classification of radar signals using the bispectrum

Features extracted from the bispectrum of radar signals are used for classification of unknown radar targets. The classification performance compared with the performance of other classifiers that are not based on higher-order spectral processing of the measured radar data. The radar signals used are experimental measurements that correspond to scattering from real radar targets. The data is corrupted with different types of disturbances that are likely to occur in a typical radar system. Although the number of data samples is relatively small and may be insufficient to produce very accurate bispectral estimates, it is concluded that the bispectrum classifier may outperform other known classifiers under conditions of colored noise and non-Gaussian noise.<<ETX>>

Applications of the bispectrum in radar signature analysis and target identification

This paper considers the classification of radar targets using features extracted from the bispectrum of the backscattered signals. The classification performance of the bispectrum- derived features is compared with that of features extracted from the impulse response and the frequency response of the unknown target. In each case, the classification performance is evaluated using compact-range radar signal measurements of a set of five commercial aircraft. A number of scenarios of noise environment and azimuth ambiguity are considered. The case where extraneous scatterers are present in the vicinity of the measured target is also examined, and the effects of the extraneous scatterers are discussed.

Characterization of ATR systems

This paper defines the ATR problem outside the boundaries of the statistical pattern recognition (SPR) problem. It is believed that the state of the art supports successful application of SPR strategies to solve recognition problems and to the extent that the automatic target recognition (ATR) problem and the SPR problem are the same, the ATR problem is quite solvable. However, ATR remains problematic is its full realization and promise and has only been solved under a set of constrained conditions - those which map into the SPR problem. These are problems where the conditions of the training set are totally representative of the conditions under the test set. The purpose of this paper is to facilitate further progress in ATR development by defining the ATR problem in a more general way that is believed to be more representative of the actual ATR problem facing various ATR users rather than the more restricted SPR definition.

Synthetic aperture radar imagery classification using deep neural networks on a neurosynaptic processor (Conference Presentation)

The abstract is not available

Machine Learning (ML) Algorithms: An overview of various techniques for target detection and classification (Conference Presentation)

The abstract is not available