String Representation of Handwritten Katakana Characters

Web Site Web Site Open Access

Bibliographic Information

Other Title
  • 手書き片仮名文字を対象としたストリング化の一手法
  • テガキ カタカナ モジ オ タイショウ トシタ ストリングカ ノ イチ シュホウ

Search this article

Abstract

In syntactical pattern recognition, a pattern must be represented by a sentence (or string) in a language. So string representation of a input pattern is important pre-processing for syntactical pattern recognition. In this paper, we propose a method of string representation for handwritten katakana characters. Each character is a continuous line pattern on a 20×20 mesh. Starting from its upper right corner, each input pattern is chain-coded cell by cell. Namely, each successive cell is coded as A, B, C or D according to its position relative to that of the current one. After three consecutive cells have been coded one of four primitives a, b, c and d, by using four concatenation relations ×, +, *, : and the parentheses ( ), the concatenation of branches, which is a string representation, completes. From some experimental results, the validity of our method is discussed.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top