FACE, IRIS, AND FINGERPRINT MULTIMODAL IDENTIFICATION SYSTEM BASED ON LOCAL BINARY PATTERN WITH VARIANCE HISTOGRAM AND COMBINED LEARNING VECTOR QUANTIZATION

• Published in: Journal of Theoretical and Applied Information Technology Vol. 89; no. 1; p. 53
• Main Authors: Shams, Mahmoud Y, Tolba, Ahmad S, Sarhan, Shahenda H
• Format: Journal Article
• Language: English
• Published: Islamabad Journal of Theoretical and Applied Information 01.07.2016
• Subjects: Binary systems (materials); Biometrics; Classifiers; Fingerprints; Histograms; Systems analysis; Variance; Vector quantization
• ISSN: 1817-3195

Believing of the importance of biometrics in this research, we have presented a fused system that depends upon multimodal biometric system traits face, iris, and fingerprint, achieving higher performance than the unimodal biometrics. The proposed system used Local Binary Pattern with Variance histogram (LBPV) for extracting the preprocessed features. Canny edge detection and Hough Circular Transform (HCT) were used in the preprocessing stage while, the Combined Learning Vector Quantization classifier (CLVQ) was used for matching and classification. Reduced feature dimensions are obtained using LBPV histograms which are the input patterns for CLVQ producing the classes as its outputs. The fusion process was performed at the decision level based on majority voting algorithm of the output classes resulting from CLVQ classifier. The experimental results indicated that the fusion of face, iris, and fingerprint has achieved higher genuine acceptance recognition rate (GAR) 99.50% with minimum elapsed time 24 sec. The evaluation process was performed using large scale subjects claiming to enter the system proving the superiority of the proposed system over the state of art.