Harry A. Schmitt

Sparse Manifold Learning with Applications to SAR Image Classification

Nonlinear data-driven dimensionality reduction techniques have recently gained popularity due to the emergence of high dimensional data sets. The algorithmic complexity and storage requirements of these techniques, however, can make them prohibitive in resource-limited applications. It is therefore beneficial to reduce the number of exemplar samples required for performing an out-of-sample extension to a test point. In this paper, we propose a novel method for selecting a minimal set of exemplars and performing the out-of-sample extension. In the case of two-class target recognition with synthetic aperture radar (SAR) data, we compare the efficacy of the proposed approach with other approaches for selecting a subset of the available training samples. We show that the proposed algorithm outperforms the existing methods by providing low-dimensional embeddings that maintain interclass separability using fewer retained exemplars.

A combined particle/Kalman filter for improved tracking of beam aspect targets

Track continuity is difficult to maintain when tracking beam aspect targets. The loss of Doppler discrimination allows clutter to mask the target return, making it nearly impossible to detect. In order to improve tracking performance, a combination particle/Kalman filter has been developed. The tracking filters obviate each other as appropriate. When a target enters a Doppler blind zone, the particle filter replaces the Kalman filter as the tracking algorithm until the target exits the zone. The particle filter expands over the uncertainty region so that when the target is once again visible, it can immediately resume track via re-initialization of the Kalman filter. This paper discusses the design and simulation of this algorithm and shows the resulting improvement in track continuity. We briefly discuss how our combined particle/Kalman filter approach can be used to address the problem of targets obscured in altitude return.

Relationships between radar ambiguity and coding theory

We investigate the theory of the finite discrete Heisenberg-Weyl group in relation to the development of adaptive radar. We contend that this group can form the basis for the representation of the radar environment in terms of operators on the space of waveforms. We also demonstrate, following recent developments in the theory of error correcting codes, that the finite discrete Heisenberg-Weyl group provides a unified basis for the construction of useful waveforms/sequences for radar, communications and the theory of error correcting codes.

Activities in integrated sensing and processing

Advances in sensor technologies, computation devices and algorithms have created enormous opportunities for intelligent, automated threat identification, target acquisition, and surveillance. Unfortunately, as information requirements grow, conventional network processing techniques require ever-increasing bandwidth between sensors and processors, as well as potentially exponentially complex data reduction methods. Practical approaches require that the sensing and computation be jointly engineered, to raise the quality of data and classification and to minimize computation, power consumption and cost. This paper summarizes several activities at Raytheon investigating the integration of sensing and processing and presents some preliminary results.

Maximum-entropy surrogation in network signal detection

Multiple-channel detection is considered in the context of a sensor network where raw data are shared only by nodes that have a common edge in the network graph. Established multiple-channel detectors, such as those based on generalized coherence or multiple coherence, use pairwise measurements from every pair of sensors in the network and are thus directly applicable only to networks whose graphs are completely connected. An approach is introduced that uses a maximum-entropy technique to formulate surrogate values for missing measurements corresponding to pairs of nodes that do not share an edge in the network graph. The broader potential merit of maximum-entropy baselines in quantifying the value of information in sensor network applications is also noted.

Asymptotic Performance of ATR in Infrared Images

In this study, the asymptotic performance analysis for target detection-identification through Bayesian hypothesis testing in infrared images is presented. In the problem, probabilistic representations in terms of Bayesian pattern-theoretic framework is used. The infrared clutter is modelled as a second-order random field. The targets are represented as rigid CAD models. Their infinite variety of pose is modelled as transformations on the templates. For the template matching in hypothesis testing, a metric distance, based on empirical covariance, is used. The asymptotic performance of ATR algorithm under this metric and Euclidian metric is compared. The receiver operating characteristic (ROC) curves indicate that using the empirical covariance metric improves the performance significantly. These curves are also compared with the curves based on analytical expressions. The analytical results predict the experimental results quite well.

Multiresolution approach to object classification using kinematic features

A multi-resolution approach to problems of the identification of classes of ballistic missile objects is outlined. This approach is based on the utilization of features estimated from time-varying infrared signatures and the subsequent discrimination of different objects using unique time-frequency patterns obtained from a multi- resolution decomposition of the training and observation (performance evaluation) data. For example, we have identified four features that show some promise for discrimination: the intensity in the second lowest sub-band, the temporal profile in the lowest frequency sub-band, the modulation intensity, and the DC level of each observed object. The multi-resolution discrimination algorithms performance can be evaluated by comparing with more traditional Fourier based approaches. The multi-resolution discrimination algorithms were applied to simulated data and were shown, by using L1 or L2 norms as distance metrics, to provide good classification performance and to reduce the temporal data length by half. The features extracted using the discrete wavelet packet transform can help to further improve classification performance. The robustness of the algorithm in the presence of noise is also studied. All data sets were generated with Raytheon Missile Systems Companys high fidelity simulation.