Open AccessIndividual biometrics pattern based artificial image analysis techniques
Author(s) -
Israa Mohammed Khudher,
Yahya Ismail Ibrahim,
Suhaib Abduljabbar Altamir
Publication year - 2021
Publication title -
numerical algebra, control and optimization
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.303
H-Index - 20
eISSN - 2155-3289
pISSN - 2155-3297
DOI - 10.3934/naco.2020056
Subject(s) - biometrics , computer science , artificial intelligence , image processing , pattern recognition (psychology) , cluster analysis , footprint , process (computing) , field (mathematics) , feature extraction , feature (linguistics) , computer vision , image (mathematics) , mathematics , paleontology , linguistics , philosophy , pure mathematics , biology , operating system
Biometric characteristics have been used since antiquated decades, particularly in the detection of crimes and investigations. The rapid development in image processing made great progress in biometric features recognition that is used in all life directions, especially when these features recognition is constructed as a computer system. The target of this research is to set up a left foot biometric system by hybridization between image processing and artificial bee colony (ABC) for feature choice that is addressed within artificial image processing. The algorithm is new because of the rare availability of hybridization algorithms in the literature of footprint recognition with the artificial bee colony assessment. The suggested system is tested on a live-captured ninety colored footprint images that composed the visual database. Then the constructed database was classified into nine clusters and normalized to be used at the advanced stages. Features database is constructed from the visual database off-line. The system starts with a comparison operation between the foot-tip image features extracted on-line and the visual database features. The outcome from this process is either a reject or an acceptance message. The results of the proposed work reflect the accuracy and integrity of the output. That is affected by the perfect choice of features as well as the use of artificial bee colony and data clustering which decreased the complexity and later raised the recognition rate to 100%. Our outcomes show the precision of our proposed procedures over others' methods in the field of biometric acknowledgment.