A Study on the Hand-written Number Recognition by HMM(Hidden Markov Model)

JO, MIN HWAN

Journal of The Korea Society of Computer and Information

Abbr : JKSCI
2004, 9(3), pp.121-125
Publisher : The Korean Society Of Computer And Information
Research Area : Computer Science

JO, MIN HWAN ¹

¹춘천교육대학교

ABSTRACT

In the most of recognizing systems of hand-written numbers, extraction of feature shape by using character elements shapes and a method of morphological analysis by using then extraction of feature shapes were usually used. In this paper, however, peculiar chain-code is used, and differential code which gets minimal value by differentiating the chain-code which is generated by the peculiar chain-code is made. We found this differential code is very successful in discriminating hand-written numbers according to the result of applying to most of the hand-written numbers. Testing recognition of hand-written numbers by HMM network. From the results, we can recognize of 96.1 percentage hand-written numbers but can not recognize extremely distorted hand-written numbers.
KEYWORDS

인식(Recognition), 방향성코드(Chain Code), HMM(Hidden Markov Model)
REFERENCES (9)

[journal] Jay June Lee / Data-driven Design of HMM Topology for On-line Handwriting Recognition / KAIST. 2000for Department of Electrical Engineering and Computer Science / Division of Computer Science

[journal] 2111-2115. Sep. 1991 / N. Merhav and Y. Ephraim. Maximun likelihood hidden Markov modeling using a dominant sequence of states. IEEE Trans. on Signal Processing

[journal] J.-Y. Ha / and Jin H. Kim. Unconstrained handwritten word recognition with interconnected hidden Markov models. In Third Int. Workshop on Frontiers in Handwriting Recognition : 455~460

[journal] Michel Gilloux / 1993. / Strategies for Handwritten Words Recognition Using Hidden Markov Models Proc. of 2nd Int. Cong. on Document Analysis and Recognition : 299~304

[journal] H. Bunke / 1995. / Off-Line Cursive Handwriting Recognition Using Hidden Markov Model 28(9) : 1399~1413

[journal] Yannick Tremblay / 1994. / Proc. of 4th Int. Workshop on Frontiers in Handwriting Recognition : 449~455

[journal] Woo Sung Kim / 1996. / Off-line Recognition of Handwritten Korean and Alphanumeric Characters Using Hidden Markov Models 29(5) : 845~858

[journal] 조성목 / 2001. / “이미지 특징 추출연산자 DBAH/DBAG 와 하드웨어 실현” 제6권 (제1호)

[journal] 조민환 / 2003. / “체인코드를 이용한 훼손된 지문의 인식에 관한 연구” ^제8권 (제4호) : 63~68
KCI Citation (1)

KJC Korea
Journal Central

Statistics

Tools