Меню

Information and communication and chemical technologies

No. 2 (23) - 2024 / 2024-06-30 / Number of views: 52

INTEGRATION OF UNIMODAL AND MULTIMODAL BIOMETRIC SYSTEMS TO IMPROVE THE ACCURACY OF PERSONAL IDENTIFICATION: DEVELOPMENT AND EVALUATION OF THE PCA-DAUGMAN METHOD

Authors

Astana International University
L. N. Gumilev Eurasian National University,
L. N. Gumilev Eurasian National University
K.Kulazhanov Kazakh University of Technology and Business
Astana International University, Astana

Keywords

biometrics, multimodal systems, face recognition, Daugman algorithm, Principal Component Analysis (PCA), which describe the main technologies and methods used to enhance the effectiveness of personal identification systems
Back

PDF (RU)

Link to DOI:

https://doi.org/10.58805/kazutb.v.2.23-488

How to quote

Nazyrova А., Mussaif М., Kintonova А., Altynbek С., and Kaldarova М. “INTEGRATION OF UNIMODAL AND MULTIMODAL BIOMETRIC SYSTEMS TO IMPROVE THE ACCURACY OF PERSONAL IDENTIFICATION: DEVELOPMENT AND EVALUATION OF THE PCA-DAUGMAN METHOD”. КазУТБ, vol. 2, no. 23, June 2024, doi:10.58805/kazutb.v.2.23-488.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver

Abstract

This article examines the application of unimodal and multimodal biometric systems for personal identification, with a particular focus on improving recognition accuracy. Unimodal systems, which use a single type of biometric feature, such as the iris or face, are widely used in areas requiring high security measures and attendance control. Despite their popularity, these systems face several challenges, including data noise and intraclass variations, which limit their effectiveness.

The paper provides a detailed analysis of various methods and algorithms used in unimodal systems, including Principal Component Analysis (PCA) for face recognition and the Daugman algorithm for iris identification. Their advantages and limitations are discussed based on recent research.

Special attention is given to the development of a multimodal biometric system that combines multiple biometric features to enhance identification accuracy. The article proposes a new combined PCA-Daugman method, integrating face and iris recognition techniques. This approach shows significant improvements in performance compared to unimodal systems, as confirmed by testing on ORL, YALE, Real face, and CASIA datasets.

The study demonstrates that a multimodal biometric system can effectively overcome the limitations of unimodal systems, offering a higher level of security and reliability in personal identification.

Kazakh university of technology and business