PATTERN-DISCRIMINATION USING A BANK OF WAVELET FILTERS IN A JOINT TRANSFORM CORRELATOR

We introduce and demonstrate the use of wavelet-filter-based processin g in a joint transform correlator for achieving high-quality target di scrimination. The inherent frequency localization characteristic of th ese filters offers high discrimination in correlation for lookalike ta rgets. Results obtained from simulation and experimental studies Valid ate this fact, Wavelet filters are generated and applied to the joint power spectrum through fast digital processing and are displayed onto a spatial light modulator, The correlation results are analyzed for ta rget discrimination. The hybrid experimental setup combines the speed of optical processing with the flexibility of digital processing to re alize such a technique in real time. (C) 1998 Society of Photo-Optical Instrumentation Engineers.