Feature-level fusion of major and minor dorsal finger knuckle patterns for person authentication
- PDF / 1,365,225 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 47 Downloads / 218 Views
ORIGINAL PAPER
Feature-level fusion of major and minor dorsal finger knuckle patterns for person authentication Abdelouahab Attia1
· Zahid Akhtar2 · Youssef Chahir3
Received: 8 August 2019 / Revised: 18 August 2020 / Accepted: 16 September 2020 © Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract The identification of individuals by their finger dorsal patterns has become a very active area of research in recent years. In this paper, we present a multimodal biometric personal identification system that combines the information extracted from the finger dorsal surface image with the major and minor knuckle pattern regions. In particular, first the features are extracted from each single region by BSIF (binarized statistical image features) technique. Then, extracted information is fused at feature level. Fusion is followed by dimensionality reduction step using PCA (principal component analysis) + LDA (linear discriminant analysis) scheme in order to improve its discriminatory power. Finally, in the matching stage, the cosine Mahalanobis distance has been employed. Experiments were conducted on publicly available database for minor and major finger knuckle images, which was collected from 503 different subjects. Reported experimental results show that feature-level fusion leads to improved performance over single modality approaches, as well as over previously proposed methods in the literature. Keywords Finger dorsal patterns · BSIF · PCA + LDA · Feature-level fusion
1 Introduction The principal idea of biometrics is replacing person recognition via human experts with machine learning-based automated identity verification system in different applications such as national identity card, driver’s license, social security, border control, passport control and mobile user authentication [1]. Broadly, biometric systems can be divided into categories: unimodal (i.e., using only single biometric source of information to establish the identity) and multibiometric system (i.e., using multiple biometric sources of information) [2]. However, the unimodal biometric systems suffer from several defects such as high error rate, low usability, the possibility of the intrusion of these systems and other problems (e.g., aging) [3, 4]. Multibiometrics is an alternative solution that merges information from multiple biometric sources
B
Abdelouahab Attia [email protected]
1
LMSE Laboratory, Mohamed El Bachir El Ibrahimi University, Bordj Bou Arreridj, Algeria
2
State University of New York Polytechnic Institute, Marcy, NY, USA
3
University of Caen, Caen, France
[5–7]. The information sources can be different instances of the same modality, different biometric modalities, numerous prototypes of same modality from different sensors or several feature extraction algorithms for the same single modality. There exist ample of theoretical and experimental studies that demonstrate the efficacy of the multimodal biometric systems that improve performance compared to unimodal systems [8–10]. Therefore, not
Data Loading...
