Online Character Recognition Technique Using PCA

PCA를 이용한 온라인 문자인식 기법

  • 유재만 (광운대학교 컴퓨터과학과) ;
  • 김우생 (광운대학교 컴퓨터공학부) ;
  • 한정훈 (광운대학교 컴퓨터과학과)
  • Published : 2006.04.01

Abstract

Online character recognition techniques have been applied in many new fields of PDA, Tablet PC etc. But the recognition techniques can not use such high technologies naturally yet. Hidden Markov Model (HMM) that is much used recently requires high memory space and complex computational tasks because of comparing the input data with entire standard patterns. In this paper we propose a method to recognize the online characters more efficiently. At first we create chain-codes of learning data and recognition data in preprocessing phase, and then we compress dimensions of data using Principal Component Analysis (PCA) and recognize a character compressed data in recognition phrase. Validity of proposed method .is verified. by experiment results.