Preprocessing techniques for cursive script word recognition

摘要

This paper deals with techniques for improving the recognition rate of a cursive script word recognition system. Closed-loop preprocessing techniques have been designed and implemented to achieve this objective on a limited vocabulary but with no restrictions on handwriting style. This paper discusses the details of such a system and its performance on samples from several authors. Results obtained from this study are promising and suggest that closed-loop verification is a potentially more useful technique than previous open-loop processing approaches.