BVCUN-10Print comprises a total of 3756 good quality ten-prints of 95 African students, acquired at four impressions per finger, in 2014. This dataset comprises images of finger types: thumb, index, middle, engagement and tiny, for the right and left fingers. Hence, at most, 40 impressions of a subject’s ten-print exist in this dataset.
Images of fingerprints were acquired using a Microsoft fingerprint reader and stored in Tagged Image File (TIF) format. This database was collected in single-session over a period of 3 months in the Department of Electronic Engineering, University of Nigeria.
The right and left fingers and stored in separate folders and the size of this database is 296.7 MB (zipped).
The fingerprints are also stored by types in sub-folders.
Fingerprints are numbered following this nomenclature Xnnn_#.tif where:
X represents the finger type and could be:
T – for thumbs
I – for index fingers
M – for middle fingers
E – for engagement finger
N – for tiny fingers
nnn stands for the subject number and runs from 000 to 095.
# represents the impression number of a subject’s image.
Since there are four impressions per subject in this database, # runs from 1 to 4. The nomenclature can be assessed from the readme file (readme_10print.txt).
In the acquisition process, a few finger types for some subjects had less than four impressions per subject. In order to maintain consistency in the number of impressions per subject’s finger type, all subjects finger types, for which their impressions that were less than four were removed.
The statistic of the total number of impressions per subject is provided in Table 1 in readme_10print.txt
Statistics of the finger types – Thumb, index, middle, engagement and tiny, for the left and right fingers are provided in Table 2 in readme_10print.txt. Number of fingerprints per finger type in Table 2 are divisible by 4.
Subjects number per finger type, that were deleted in the collection are provided in Table 3, per finger type in this file.
A list containing subjects’ numbers, whose four fingerprints exist in all ten prints sub-folders are provided in the file, common_identifiers_subjects_list_numeric.txt (for a numeric version) or common_identifiers_subjects_list_strings.txt (for strings version). This list provides the 85 subjects numbers whose complete 40 impressions of 10 fingerprint types exist in the database.
When the BVCUN-10Print dataset is used in any form of research, please the following paper should be cited:
Ogechukwu N. Iloanusi,
Fusion of finger types for fingerprint indexing using minutiae quadruplets,
Pattern Recognition Letters,
√ BVCUN-10Print dataset (BVCUN-10Print.zip)
√ Readme file (readme_10print.txt)