Hybrid optical correlator for character recognition

A hybrid character recognition system is described. It is composed of two parts, the feature extractor and the inner-product correlator. Ten Arabic characters from 0 to 9 are tested for their recognition by the system. The experimental results show that all the printing charactersare perfectly recognized with rotational invariance of 15° angle, and for hand-writing characterstheir recognized ratio is over sixty percent. 1. INTRODUCTION For the consideration of rapidly image identifying and pattern recognizing, the researchon image processing by various optical processors and hybrid processors has become an active subject.' —6 The techniques of feature extraction, optical correlation and artificial neural network, we think, are most useful for image processing and pattern recognition. In practice, however, the number of object class that requires identification is large and an object includes ahuge information. So that only use of one image processing technique always bears beyond it'sability for object recognition, and a combination of several such techniques has developed forimage processing and object recognition. Such three optical information processing techniques7have been suggested to be combined to achieve the best benefits in image processing.