A new hierarchical approach for recognition of unconstrained handwritten numerals

Author

Wang, Gwo-En ; Wang, Jhing-Fa

Author_Institution

Dept. of Electr. Eng., Nat. Cheng Kung Univ., Tainan, Taiwan

Volume

40

Issue

3

fYear

1994

fDate

8/1/1994 12:00:00 AM

Firstpage

428

Lastpage

436

Abstract

A new hierarchical approach for the recognition of unconstrained handwritten numerals is proposed. In order to obtain a reliable skeleton of the observed character, some preprocessing operations including smoothing, noise removal, normalization, and a thinning process are first applied to each character. Then, some interesting feature points are extracted from this reliable skeleton of the character. In the first stage of preclassification, a set of structural features named four-zone codes is adopted to preclassify the numerals. Due to the large degree of data and distortion of characters, a three layer fuzzy neural network is used for fine classification. Experimental results show that a high recognition rate over 99.5% is obtained

Keywords

feature extraction; fuzzy set theory; image coding; image recognition; neural nets; feature points; fine classification; four-zone codes; hierarchical approach; noise removal; normalization; preclassification; preprocessing; recognition; recognition rate; reliable skeleton; smoothing; structural features; thinning; three layer fuzzy neural network; unconstrained handwritten numerals; Character recognition; Data mining; Feature extraction; Fuzzy neural networks; Handwriting recognition; Neural networks; Pattern recognition; Reliability engineering; Shape; Skeleton;

fLanguage

English

Journal_Title

Consumer Electronics, IEEE Transactions on

Publisher

ieee

ISSN

0098-3063

Type

jour

DOI

10.1109/30.320824

Filename

320824