A handwritten Japanese historical Kana reprint Support System based on pattern recognition and constraint solving
Project/Area Number |
16K00463
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Library and information science/Humanistic social informatics
|
Research Institution | Shibaura Institute of Technology |
Principal Investigator |
SUZUKI Tetsuya 芝浦工業大学, システム理工学部, 教授 (00323824)
|
Co-Investigator(Kenkyū-buntansha) |
相場 亮 芝浦工業大学, システム理工学部, 教授 (40317368)
|
Project Period (FY) |
2016-04-01 – 2019-03-31
|
Project Status |
Completed (Fiscal Year 2018)
|
Budget Amount *help |
¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000)
Fiscal Year 2018: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2017: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2016: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
|
Keywords | 歴史情報 / 翻刻支援 / 変体仮名 / 制約プログラミング / 文字認識 |
Outline of Final Research Achievements |
The purpose of this research is to realize reprint of handwritten Japanese historical Kana text by combining pattern recognition and constraint solving, which was not achieved by the conventional methods. The system which our research aims selects appropriate character recognition results as a whole from combination of plural character recognition results of each character using a constraint solver. We implemented the constraint solver as a Web service and a system with a graphical user interface (GUI) which invokes the constraint solver. The system makes it possible to generate a constraint satisfaction problem (CSP) from character segmentation results, character reading order, and plural recognition results of each character specified via the GUI and to show solutions to the CSP visually.
|
Academic Significance and Societal Importance of the Research Achievements |
本研究は古文書の翻刻作業を支援する技術に関するものである。古文書には現行仮名だけでなくその異体字である変体仮名も使われている。さらにそれらが崩された形で書かれている。そのため現代の日本人が古文書を読むことが困難になっている。本研究の特徴は、全体として自然な文となるように一文字ずつの文字認識結果を補正する点にある。このような技術によって古文書の内容がコンピュータで活用しやすくなる。
|
Report
(4 results)
Research Products
(6 results)