Open AccessPersonal identification system based on multi biometric depending on cuckoo search algorithm
Author(s) -
Ansam Nazar Younis
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1879/2/022080
Subject(s) - biometrics , computer science , identification (biology) , cuckoo search , histogram , face (sociological concept) , artificial intelligence , identity (music) , pattern recognition (psychology) , image (mathematics) , face detection , computer vision , facial recognition system , algorithm , social science , botany , physics , particle swarm optimization , sociology , acoustics , biology
In modern devices, many personal identification systems are used using various biometrics to confirm the identity of an individual and identify him for several purposes. Some of these essential biometrics are used in this paper to help identify a person while attaining social distance because of the widespread epidemics. The features of the face, eyes, nose, and finally, the features of the mouth are used in this article. The work begins by detecting the parts of multibiometric from the input images using Viola-Jones face detection algorithm with a modification to it then segment them. After that, various initial treatment processes begin which helps clarify them to facilitate subsequent operations. Also, a Histogram of oriented gradient method(HOG) is used to extract the significant features of those image segments. The extracted features from these segments are entered into the developed cuckoo search algorithm(DCSA), the best image segment within the used dataset is searched for similar in terms of characteristics to the entered image segment. The work has also been developed so that the system is executed on two cores using parallel processing technology to utilize the processor as much as possible and reduce the time it takes to implement the system and identify the person concerned. A high identification rate has been reached, reaching 99.25%, and overall speed up 1.40323 sec relative to serial execution.