Problems in Chinese Character Recognition, Proceedings of 1st USA_JAPAN Computer Conference, pp.296-301 (1972)
Introduction
The Japanese language is normally written, typed, or printed in a mixture of
Chinese ideographs and Japanese syllabic (kana) characters.
Recognition of these Chinese characters is expected to be a bottleneck of computer
use in business, management, and many other nonscientific fields in Japan.
A recognition of Chinese characters creates two new problems concerning the OCR
--- (1) how to manage the large size of the character set as indicated in table
1 and (2) how to handle complex patterns such as those shown in Fig. 1.
Chinese characters can be classified into four categories from a
recognition viewpoint --- printed (or typed) single font, printed multifont,
handwritten, and on-line handwriting.
In this article, some approaches ro the
problems are briefly described in the case of Chinese character recognition of
the Educational Set which contains 881 characters.
| mail address: | |
← お手数ですが打ち込んで下さい |
First Written March 23, 2000
Transplanted to KSU Before May 17, 2003
Transplanted to So-net May 3, 2005
Last Update April 22, 2007
© Yasuaki Nakano 2000-2007