Fuzzy technique based recognition of handwritten characters

The different methods for automatic pattern recognition are motivated by the way in which pattern classes are characterized and defined. In this paper, the handwritten characters (numerals) are preprocessed and segmented into primitives. These primitives are measured and labeled using fuzzy logic. Strings of a character are formed from these labeled primitives. To recognize the handwritten characters, conventional string matching is performed. However, the problem in this string matching has been avoided using the membership value of the string. This result is being compared with the Modified Parser generated from the Error-free fuzzy context-free grammar.