Research on Automatic Handwritten-like Font Creation that Reflects Individual Writing Habit
Project/Area Number |
13650423
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
情報通信工学
|
Research Institution | Keio University |
Principal Investigator |
NAKAJIMA Masato Keio University, Dept.of Electronics and Electrical Engineering, Professor, 理工学部, 教授 (20051766)
|
Co-Investigator(Kenkyū-buntansha) |
SAITO Hideo Keio University, Dept.of Information Science and Technology, Associate Professor, 理工学部, 助教授 (90245605)
|
Project Period (FY) |
2001 – 2002
|
Project Status |
Completed (Fiscal Year 2002)
|
Budget Amount *help |
¥2,300,000 (Direct Cost: ¥2,300,000)
Fiscal Year 2002: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 2001: ¥1,500,000 (Direct Cost: ¥1,500,000)
|
Keywords | Image Processing / Pattern Recognition / Individuality / Individual Writing Habit / Character / Font / Standard Character / Correspondence Estimation / 個人のくせ / 文字画像処理 / 特徴抽出 / エネルギー汎関数 |
Research Abstract |
Everyone can write beautiful characters because of the diffusion of word processor in recent years. But on the other side, there are people who are not satisfied with uniformly font, too. Some warmth will be able to be added to communications if a character font that reflects individual writing habit can be generated. In the source of such thought, we proposed a method which generates an arbitrary character font that reflects individual writing habit automatically. At first, a character group that an user writes is inputted by image scanner. Handwritten-like character fonts which are the standard fonts has already been installed. Then, the computer analyzes whether geometrical transformation and scale conversion are done between the standard characters font and the enter characters, and the user's individual writing habit is extracted. If the character transformation rule is applied conversely toward the standard character font, arbitrary kind of character that reflects individual writing habit can be generated automatically. We developed a system called "writing habit including character font generator" which includes the proposed algorithm, and its performance was evaluated. The system indicates the device which can extract individual writing habit from handwritten characters which are inputted by the image scanner and can express on display or print out the characters that reflect the individual writing habit when character codes are entered with keyboard. The effectiveness of the algorithm was confirmed by the quantitative examination and the comparison by the visual observation toward characters that many subjects write and outputted character fonts that reflect the subjects' individual writing habit. The result of the study was introduced abundantly to the paper, journal and so on. Furthermore, the result was transfered to a company at once, and the sale of the software product is started in February 2003.
|
Report
(3 results)
Research Products
(7 results)