A Study on Support System for Visually Handicapped People by Extracting Information from Scene Images
| Project/Area Number |
16K01547
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Rehabilitation science/Welfare engineering
|
|---|
| Research Institution | Kobe University |
|---|
Principal Investigator |
Tanaka Naoki 神戸大学, 海事科学研究科, 教授 (90188318)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
鎌原 淳三 神戸大学, 海事科学研究科, 准教授 (60283917)
長松 隆 神戸大学, 海事科学研究科, 准教授 (80314251)
熊澤 宏之 大阪産業大学, 工学部, 教授 (00592320)
|
|---|
| Project Period (FY) |
2016-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥4,810,000 (Direct Cost: ¥3,700,000、Indirect Cost: ¥1,110,000)
Fiscal Year 2019: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2018: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2017: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2016: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
|
|---|
| Keywords | 視覚障がい者 / 情景画像 / オブジェクト抽出 / 歩行支援 / ディープラーニング / 点字ブロック / 文字列抽出 / 視覚障がい / CNN / FCN / エッジ強調 / セグメンテーション |
|---|
| Outline of Final Research Achievements |
People with visual disabilities are seeing an increase in the number of accidents involving falls from railway station platforms, and are required to provide walking support to ensure their safety and expand their range of activity. In this research project, we conducted a study on the extraction of walking support objects for visually impaired persons such as character strings such as signboards and Braille blocks from the scene images acquired by wearable devices, and the classification of them. We have achieved achievements such as low-resolution character string enhancement method, character string area extraction, Braille block discrimination, and area extraction method development.
|
| Academic Significance and Societal Importance of the Research Achievements |
歩行支援オブジェクトの抽出方式には、主に、ディープラーニングを用いるが、歩行支援を実現するためには、案内標識の文字列や、点字ブロックの領域を抽出しその意味内容、レイアウト構造の解析が必要となる。そのための、学習データの作成方法や、領域抽出・解析方式の開発はディープラーニングの更なる活用に繋がる意味で学術的に意義がある。また、視覚障がい者の歩行支援は社会的なニーズも高いものと考えられる。
|
Report
(5 results)
Research Products
(9 results)
