DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER
For improved human comprehension and autonomous machine perception, optical character recognition has been saddled with the task of translating printed or hand-written materials into digital text files. Many works have been proposed and implemented in the computerization of different human language...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Alma Mater Publishing House "Vasile Alecsandri" University of Bacau
2021-12-01
|
| Series: | Journal of Engineering Studies and Research |
| Subjects: | |
| Online Access: | https://jesr.ub.ro/index.php/1/article/view/302 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823858334886789120 |
|---|---|
| author | OLUWASHINA O. OYENIRAN JOSHUA O. OYENIYI LAWRENCE O. OMOTOSHO IBRAHIM K. OGUNDOYIN |
| author_facet | OLUWASHINA O. OYENIRAN JOSHUA O. OYENIYI LAWRENCE O. OMOTOSHO IBRAHIM K. OGUNDOYIN |
| author_sort | OLUWASHINA O. OYENIRAN |
| collection | DOAJ |
| description |
For improved human comprehension and autonomous machine perception, optical character recognition has been saddled with the task of translating printed or hand-written materials into digital text files. Many works have been proposed and implemented in the computerization of different human languages in the global community, but microscopic attempts have also been made to place Yoruba Handwritten Character on the board of Optical Character Recognition. This study developed a novel available dataset for research on offline Yoruba handwritten character recognition so as to fill the gaps in the existing knowledge. The developed database contains a total of 12,600 characters being made up of 70 classes from a total number of 200 writers, in which 80 % (10,500) is regarded as the training and validation dataset while the remaining 20 % (2,100) is regarded as testing dataset. The dataset is available on https://github.com/oluwashina90/Yoruba-handwritten-character-database. Hence, it is the complete and largest dataset available for Yoruba Handwritten character research.
|
| format | Article |
| id | doaj-art-4ffb1084eeba4c9bbf5127c7558d78f0 |
| institution | Kabale University |
| issn | 2068-7559 2344-4932 |
| language | English |
| publishDate | 2021-12-01 |
| publisher | Alma Mater Publishing House "Vasile Alecsandri" University of Bacau |
| record_format | Article |
| series | Journal of Engineering Studies and Research |
| spelling | doaj-art-4ffb1084eeba4c9bbf5127c7558d78f02025-02-11T11:40:02ZengAlma Mater Publishing House "Vasile Alecsandri" University of BacauJournal of Engineering Studies and Research2068-75592344-49322021-12-0127410.29081/jesr.v27i4.302DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTEROLUWASHINA O. OYENIRANJOSHUA O. OYENIYILAWRENCE O. OMOTOSHOIBRAHIM K. OGUNDOYIN For improved human comprehension and autonomous machine perception, optical character recognition has been saddled with the task of translating printed or hand-written materials into digital text files. Many works have been proposed and implemented in the computerization of different human languages in the global community, but microscopic attempts have also been made to place Yoruba Handwritten Character on the board of Optical Character Recognition. This study developed a novel available dataset for research on offline Yoruba handwritten character recognition so as to fill the gaps in the existing knowledge. The developed database contains a total of 12,600 characters being made up of 70 classes from a total number of 200 writers, in which 80 % (10,500) is regarded as the training and validation dataset while the remaining 20 % (2,100) is regarded as testing dataset. The dataset is available on https://github.com/oluwashina90/Yoruba-handwritten-character-database. Hence, it is the complete and largest dataset available for Yoruba Handwritten character research. https://jesr.ub.ro/index.php/1/article/view/302database, yoruba, handwritten character |
| spellingShingle | OLUWASHINA O. OYENIRAN JOSHUA O. OYENIYI LAWRENCE O. OMOTOSHO IBRAHIM K. OGUNDOYIN DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER Journal of Engineering Studies and Research database, yoruba, handwritten character |
| title | DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER |
| title_full | DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER |
| title_fullStr | DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER |
| title_full_unstemmed | DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER |
| title_short | DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER |
| title_sort | development of an improved database for yoruba handwritten character |
| topic | database, yoruba, handwritten character |
| url | https://jesr.ub.ro/index.php/1/article/view/302 |
| work_keys_str_mv | AT oluwashinaooyeniran developmentofanimproveddatabaseforyorubahandwrittencharacter AT joshuaooyeniyi developmentofanimproveddatabaseforyorubahandwrittencharacter AT lawrenceoomotosho developmentofanimproveddatabaseforyorubahandwrittencharacter AT ibrahimkogundoyin developmentofanimproveddatabaseforyorubahandwrittencharacter |