Effects of Polarization and Resolution on the Performance of a SAR Automatic Target Recognition System

■ Lincoln Laboratory is investigating the use of high-resolution fully polarimetric synthetic-aperture radar (SAR) imagery to detect and classify stationary ground targets. This article summarizes a study in which data collected by the Lincoln Laboratory 33-GHz SAR were used to perform a comprehensive comparison of automatic target recognition (ATR) performance for several combinations of polarization and resolution. The Lincoln Laboratory baseline ATR algorithm suite was used, and it was optimized for each polarization/resolution case. The polarizations evaluated were HH and PWF; the resolutions evaluated were 1 ft × 1 ft and 1 m × 1 m. The data set used for this study contained approximately 74 km of clutter (56 km of mixed clutter and 18 km of highly cultural clutter) and 136 tactical target images divided equally between tanks and howitzers.

[1]  Richard O. Duda,et al.  Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.

[2]  G. B. Goldstein,et al.  False-Alarm Regulation in Log-Normal and Weibull Clutter , 1973, IEEE Transactions on Aerospace and Electronic Systems.

[3]  K. Fukunaga,et al.  The Acquisition Probability for a Minimum Distance One-Class Classifier , 1987, IEEE Transactions on Aerospace and Electronic Systems.

[4]  Leslie M. Novak,et al.  Texture discrimination in synthetic aperture radar imagery , 1989, Twenty-Third Asilomar Conference on Signals, Systems and Computers, 1989..

[5]  R. D. Chaney,et al.  Optimal Processing of Polarimetric Synthetic-Aperture Radar Imagery , 1990 .

[6]  Leslie M. Novak,et al.  Optimal polarimetric processing for enhanced target detection , 1993 .

[7]  Christine M. Netishen,et al.  Performance of a High-Resolution Polarimetric SAR Automatic Target Recognition System , 1993 .