OPTODIGITAL NEURAL-NETWORK CLASSIFIER

A two-layer neural network architecture for carrying out optodigital c lassification operations is proposed. The optical neural network imple mentation is suitable for pattern recognition and classification into digital format. The neural network is based on an optical correlator, an optoelectronic threshold, and an optodigital encoder. The module ne eds only one laser light source, and the light propagation from the in put to the output is uninterrupted. The output is the class of the inp ut pattern encoded in a digital format. Experimental results using dif ferent images and classes are presented. The classification can be cha nged arbitrarily by changing the encoding mask. (C) 1997 Society of Ph oto-Optical instrumentation Engineers.