Recognizing articulated objects in SAR images

This paper presents the first sucessful approach for recognizing articulate d vehicles in real synthetic aperture radar (SAR) images. This approach is based on invariant properties of the objects. Using SAR scattering center l ocations and magnitudes as features, the invariance of these features with articulation (e.g. turret rotation of a tank) is shown for XPATCH-generated synthetic SAR signatures and actual signatures from the MSTAR (public) dat a. Although related to geometric hashing, our recognition approach is speci fically designed for SAR, taking into account the great azimuthal variation and moderate articulation invariance of SAR signatures. We present a basic recognition system for the XPATCH data, using scatterer relative locations , and an improved recognition system, using scatterer locations and magnitu des, that achieves excellent results with the more limited articulation inv ariance encountered with the real SAR targets in the MSTAR data. The articu lation invariant properties of the objects are used to characterize recogni tion system performance in terms of probability of correct identification a s a function of percent invariance with articulation. (C) 2000 Pattern Reco gnition Society. Published by Elsevier Science Ltd. All rights reserved.