RECOGNITION OF HAND-PRINTED LATIN CHARACTERS BASED ON A STRUCTURAL APPROACH WITH A NEURAL-NETWORK CLASSIFIER

Character recognition systems can contribute tremendously to the advan cement of the automation process and can improve the interaction betwe en person and machine in many applications, including office automatio n, check verification, and a large variety of banking, business, and d ata entry applications. Our main theme is the automatic recognition of hand-printed Latin characters using artificial neural networks in com bination with conventional techniques. This approach has a number of a dvantages. it combines rule-based (structural) and classification test s; it is more efficient for large and complex sets; and feature extrac tion is inexpensive and execution time is independent of handwriting s tyle and size. The technique can be divided into three major steps. Th e first step is preprocessing in which the original image is transform ed into a binary image utilizing a 300 dpi scanner and then thinned us ing a parallel thinning algorithm. Second, the image skeleton is trace d from left to right to build a binary tree. Some primitives, such as straight lines, curves, and loops are extracted from the binary tree. Finally, a three layer artificial neural network is used for character classification. The system was tested on a sample of handwritten char acters from several individuals whose writing ranged from acceptable t o poor in quality and the correct recognition rate obtained was 91%. ( C) 1997 SPIE and IS&T.