SYNTHESIZED MULTILEVEL COMPOSITE FILTER FOR SYNTHETIC-APERTURE RADAR IMAGE IDENTIFICATION

A novel optical correlator system to identify the Synthetic-Aperture R adar (SAR) image based on multilevel composite function (MCF) filters, constructed in input domain, is presented. The MCFs are synthesized b y the simulated annealing algorithm, which are suitable to current spa tial light modulators because of their limited dynamic range requireme nt of the MCFs. In this paper, we focus our investigation on the MCFs for the SAR image identification, including the impact of number of le vels on discriminability and noise performance. In addition, a unique position encoding technique is also added in the optical system so tha t MCF fan be implemented easily. The simulated and experimental result s are presented, which are quite consistent. (C) 1998 Elsevier Science B.V.