<title>MSTAR's extensible search engine and model-based inferencing toolkit</title>

Abstract

The DARPA/AFRL "Moving and Stationary Target Acquisition and Recognition" (MSTAR) program is developing a modelbased vision approach to Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR). The motivation for this work is to develop a high performance ATR capability that can identify ground targets in highly unconstrained imaging scenarios that include variable image acquisition geometry, arbitrary target pose and configuration state, differences in target deployment situation, and strong mtra-class variations. The MSTAR approach utilizes radar scattering models in an on-line hypothesize-and-test operation that compares predicted target signature statistics with features extracted from image data in an attempt to determine a "best fit" explanation of the observed image. Central to this processing paradigm is the Search algorithm, which provides intelligent control in selecting features to measure and hypotheses to test, as well as in making the decision about when to stop processing and report a specific target type or clutter. Intelligent management of computation performed by the Search module is a key enabler to scaling the model-based approach to the large hypothesis spaces typical ofrealistic ATR problems. In this paper, we describe the present state of design and implementation ofthe MSTAR Search engine, as it has matured over the last three years of the MSTAR program. The evolution has been driven by a continually expanding problem domain that now includes 30 target types, viewed under arbitrary squint/depression, with articulations, reconfigurations, revetments, variable background, and up to 30% blocking occlusion. We believe that the research directions that have been inspired by MSTAR's challenging problem domain are leading to broadly applicable search methodologies that are relevant to computer vision systems in many areas.

<title>Some results from scattering-based tomography for HRR and SAR prediction</title>

“…Previous studies using SAR data have demonstrated the existence of persistent scatterers 18 . These scatterers reflect radiation back to the sensor over an extended range of view angles.…”

Score-based SAR ATR performance model with operating condition dependencies

“…Some form of confidence indication is especially important for machine consumption of ATR products, for example by a fusion system. This particular ATR modeling effort was part of a study of decision-level fusion algorithms (the AFRL/SNA Fusion for Identifying Targets Experiment (FITE) [7,8]). Currently, FITE supports ATR models for two types of ATR algorithms, and these are SAR-based (this article) and IR-based [9].…”

Recognizing occluded objects in SAR images

“…There are no real SAR data with occluded objects available to the general public (limited data on vehicles in revetments [14] and partially hidden behind walls [19] has been reported to exist, but it has not yet been released for unrestricted use). In addition, there is no standard, accepted method for characterizing or simulating occluded targets.…”

“…Some of the SAR recognition techniques, e.g., MINACE filters [6], PERFORM [10], mean squared error template matching [14] and invariant histograms [9] have reported limited test results for small amounts of occlusion, typically 25 percent or less. In addition, the developers of the MSTAR search engine reported [19] using a shadow inferencing technique to hypothesize targets with up to 30 percent occlusion in the cross-range direction.…”