Analysis of Human Recognition Mechanism for Individuality and Character Variation of Hand-Written Characters and its Application to Generating Hand-Written Style Characters
| Project/Area Number |
11680212
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Educational technology
|
|---|
| Research Institution | YOKOHAMA NATIONAL UNIVERSITY |
|---|
Principal Investigator |
TAJIMA Fumiaki Yokohama National University, Dept.Edu.and Human Scis., Associate Professor, 教育人間科学部, 助教授 (10236523)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MIYATAKE Naoki Hosei University, Dept.Measurement Eng., Lecturer, 工学部, 専任講師 (40130752)
|
|---|
| Project Period (FY) |
1999 – 2000
|
| Project Status |
Completed (Fiscal Year 2000)
|
| Budget Amount *help |
¥2,100,000 (Direct Cost: ¥2,100,000)
Fiscal Year 2000: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 1999: ¥1,400,000 (Direct Cost: ¥1,400,000)
|
|---|
| Keywords | hand-written style / individuality / character variation / character generation / directional line-element / similarity / entropy / fuzzy rule |
|---|
| Research Abstract |
Summary of first study : A character learning system needs having the function of evaluating hand-written characters written by learners. We proposes a method of introducing gradation processing and Directional Line-element Entropy (DLE) into evaluation of hand-written character patterns.
The variation values, which were subjectively evaluated by some subjects for sample characters, were not able to agree with the ones evaluated by the previous method in the case of the samples with small variation. But they were able to agree with the ones by the proposed method. The variation values for the samples can be presumed with only about 30 characters.
Summary of second study : If the document can be written with hand-written style fonts in computer education, it has a good effect on pupils. After user's hand-written characters were collected on every category of characters, each stroke of the characters was converted into the time series pattern of directional line-element. We experimentally generated hand-written style character sets with character variation, based on the distribution of the weight obtained by the gradation processing using fuzzy rules. Then the variation of the generated characters about the position of their line-element was larger than that about the direction of their line-element.
|
Report
(3 results)
Research Products
(14 results)
