High accuracy handwritten Chinese character recognition by improved feature matching method

Author

Liu, Cheng-Lin ; In-Jung Eim ; Kim, Jin H.

Author_Institution

Dept. of Comput. Sci., Korea Adv. Inst. of Sci. & Technol., Taejon, South Korea

Volume

2

fYear

1997

fDate

18-20 Aug 1997

Firstpage

1033

Abstract

Proposes some strategies to improve the recognition performance of a feature matching method for handwritten Chinese character recognition (HCCR). Favorable modifications are given to all stages throughout the recognition. In pre-processing, we devised a modified nonlinear normalization algorithm and a connectivity-preserving smoothing algorithm. For feature extraction, an efficient directional decomposition algorithm and a systematic approach to design a blurring mask are presented. Finally, a modified LVQ3 algorithm is applied to optimize the reference vectors for classification. The integrated effect of these strategies significantly improves the recognition performance. Recognition results on the large-vocabulary databases ETL8B2 and ETL9B are promising

Keywords

feature extraction; handwriting recognition; image classification; image matching; optical character recognition; smoothing methods; ETL8B2 database; ETL9B database; LVQ3 algorithm; blurring mask design; classification; connectivity-preserving smoothing algorithm; directional decomposition algorithm; feature extraction; feature matching method; high-accuracy handwritten Chinese character recognition; large-vocabulary databases; learning vector quantization; nonlinear normalization algorithm; pre-processing; recognition performance; reference vector optimization; Algorithm design and analysis; Artificial intelligence; Character recognition; Computer science; Feature extraction; Handwriting recognition; Shape; Smoothing methods; Vocabulary; Writing;

fLanguage

English

Publisher

ieee

Conference_Titel

Document Analysis and Recognition, 1997., Proceedings of the Fourth International Conference on

Conference_Location

Ulm

Print_ISBN

0-8186-7898-4

Type

conf

DOI

10.1109/ICDAR.1997.620666

Filename

620666