Character Recognition Using Attributed Grammar

초록

We describe a character recognition algorithm for Korean characters by employing attribute dependent regular programmed grammar (ADRPG). Each Korean character has more strokes than a English character, and the recognition of Korean characters is heavily dependent on the attributes of strokes. We implemented a syntactic algorithm based on the ADPRG, and the ADRPG can easily incorporates attribute information in character pattern, such as position, angle, length, and branch points to recognize characters with many strokes. The algorithm based on the ADPRG is implemented on a PC-AT compatible, and 377 production rules and 78 attribute test functions are derived from a printed document containing 4206 Korean characters. The productions and the attribute test functions obtained from the 4206 characters are applied to another document containing 3092 Korean characters having some defects. More than 95.1% of 3092 Korean characters are correctly recognized by the algorithm based on the ADPRG with average speed of 1.15 characters/second.