2004 Fiscal Year Final Research Report Summary
Study on High Accuracy for Historical Character Recognition
| Project/Area Number |
14380184
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
情報システム学(含情報図書館学)
|
|---|
| Research Institution | Kyoto University (2003-2004)
International Research Center for Japanese Studies (2002) |
|---|
Principal Investigator |
SHIBAYAMA Mamoru Kyoto University, Center for Southeast Asian Studies, Professor, 東南アジア研究所, 教授 (10162645)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KATO Nei Tohoku University, Graduate School of Informatics, Professor, 大学院・情報科学研究科, 教授 (00236168)
YAMADA Shoji International Research Center for Japanese Studies, Research Division, Associate Professor, 研究部, 助教授 (20248751)
KAWAGUCHI Hiroshi Tezukayama University, Faculty of Information and Management, Professor, 経営情報学部, 教授 (80224749)
HRA Shoichiro National Institute of Japanese Literature, Associate Professor, 研究情報部, 助教授 (50218616)
NAMIKI Mitaro Tokyo University of Agriculture and Technology, Faculty of Engineering, Associate Professor, 工学部, 助教授 (10208077)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Keywords | Ancient Document / Character Recognition / Handwritten OCR / Character Recognition Dictionary |
|---|
| Research Abstract |
The aim of this research is to study a high precise recognizing system for historical character recognition by applying the hand written OCR technology, in HCR project. During the research period, the building of historical character database and reconsidering for applying character recognition technology to the historical document with writing brush have been carried out as follows.
(1)Character segmentation and normalization : Projection histogram and method of character shape have been studied. And, the column extraction based on the Hough transformation was revised and experimented.
(2)Offline character recognition : the coordination between new segmentation method for historical document with writing brush and its recognition was considered.
(3)Online character recognition : new online recognizing method for historical characters was studied based on the tablet input which would be effective for retrieving "Kuzushi" character. The electronic dictionary which can be used on Windows PC will be published in 2005.
(4)A online retrieval software was developed based on the database of "Kanji Kuzushi-ji Jiten" published by Tokyo-do Publishing Company.
(5)As for the character pattern dictionary for recognizing historical characters, 9 databases were released for the public use which can be downloaded from Web site. Homepage for HCR project is http://www.nichibun.ac.jp/shoji/hcr. Also, 2 software programs were opened, GetMoji macro and GetMoji for Web respectively.
|
