<title>Optoelectronic techniques for real-time pattern recognition</title>

Abstract

We investigate the recent trends and advancements in real time pattern recognition with special reference to joint transform correlator (JTC) architectures and algorithms. Techniques for achieving both single and multiple target detection in noise-free, noisy and cluttered input scenes using the classical, binary and fringe-adjusted JTCs are discussed. Distortion invariant target recognition using tools such as the synthetic discriminant function is also incorporated. Various optoelectronic and all-optical architectures and implementation issues for effectively implementing the various types of JTC techniques are discussed. Extensive simulation results are presented to investigate the effectiveness of the various JTC techniques, especially the fringe-adjusted JTC.