| Title |
AFHIRIS: African human iris dataset (version 1) / |
| Authors |
Akande, Oluwatobi ; Ojimba, Nzube ; Oghenekaro, Atele ; Abikoye, Oluwakemi ; Ogundokun, Roseline ; Akindele, Akinyinka |
| DOI |
10.12688/f1000research.122759.1 |
| Full Text |
|
| Is Part of |
F1000Research.. London : F1000 Research Ltd. 2022, vol. 11, art. no. 1549, p. 1-10.. ISSN 2046-1402. eISSN 1759-796X |
| Keywords [eng] |
African human iris images ; age prediction ; biometrics ; ethnicity prediction ; gender prediction ; personal recognition |
| Abstract [eng] |
Biometric systems remain the most widely used methods for identification and authentication purposes. Their wide acceptability has opened up more research into new application areas of biometric systems. However, biometric research requires an appropriate biometric dataset to validate the proposed technique. This dataset could be privately owned or publicly available for research purposes. In the field of iris biometric research, the iris dataset produced by the Chinese Academy of Sciences (CASIA) is the first, most popular, and widely used publicly available iris dataset. However, the increasing popularity and acceptability of human iris-related research have called for additional benchmarks, and therefore, new publicly available databases of human iris images. Existing publicly available human iris datasets have been collected from non-African subjects; therefore, this dataset is the first publicly available human iris dataset of African descent. Three categories of images were collected from 1028 volunteers that participated in the data collection task. The first category was made up of four iris images that were captured when the volunteers used spectacles, while the second category includes four iris images that were captured when the volunteers wore no spectacles. However, the third category of iris images was obtained from eight volunteers that used print-patterned contact lenses. Only four images were captured from volunteers in this category as they were not asked to put on spectacles. In addition to the iris images captured, soft biometric features such as age, gender, state of origin, weight, and height of the volunteers were also captured. It is strongly believed that this unique collection of iris datasets of African descent will open up new research in the study of the human iris. |
| Published |
London : F1000 Research Ltd |
| Type |
Journal article |
| Language |
English |
| Publication date |
2022 |
| CC license |
|
