| Project/Area Number |
25330217
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Perceptual information processing
|
|---|
| Research Institution | Tokyo National College of Technology |
|---|
Principal Investigator |
SUZUKI MASATO 東京工業高等専門学校, その他部局等, 教授 (50290721)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
MATSUMOTO AKIYO 東北学院大学, 教養学部, 准教授 (40413752)
|
|---|
| Co-Investigator(Renkei-kenkyūsha) |
KITAKOSHI DAISUKE 東京工業高等専門学校, 情報工学科, 准教授 (50378238)
|
|---|
| Project Period (FY) |
2013-04-01 – 2016-03-31
|
| Project Status |
Completed (Fiscal Year 2015)
|
| Budget Amount *help |
¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2014: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2013: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Keywords | 線形結合歪度モデル / パタン認識 / データマイニング / 自然言語処理 |
|---|
| Outline of Final Research Achievements |
Because of the variance of character shapes, recognition accuracy is very low for the low quality handwritten Japanese text we usually write down as a reminder. To solve this problem, we proposed the new algorithm, which extracts the components of characteristic features with own method - skewness component analysis, and which constructs the discriminant function using the linear combination of the partial normal distributions. And, we proposed a post-processing method of the character recognition using a linear combination model for the feature distributions and natural language processing. Furthermore, we reduced the computation time of these algorithms utilizing the subspaces which are used in the most nearest search algorithm (IMI). In the recognition experiments using handwritten text, we have realized a high recognition accuracy of 84.5%.
|
