MULTIMODAL FINGER DORSAL KNUCKLE MAJOR AND MINOR PRINT RECOGNITION SYSTEM BASED ON PCANET DEEP LEARNING

Abstract
Hand-based recognition systems with different traits are widely used techniques and are trustworthy ones. We can find it in different real life fields such as banks and industries due to its stability, reliability, acceptability, and the wide range features. In this paper, we present a finger dorsal knuckle print multimodal recognition system, where we use PCAnet (principal component analysis network) deep learning to extract the features from both Major and Minor finger dorsal knuckles to allow a deeper insight into the exploited trait. Then SVM is used for the matching stage of the two modalities, followed by a score level fusion method to combine the scores using different rules. In order to establish the effectiveness of the proposed system, several experiments in the course of this work have been done on the finger knuckle images of the publicly available database PolyUKV1. The results show that the proposed method has better results in comparison with a unimodal system.

Authors
Nour Elhouda Chalabi1, Abdelouahab Attia2, Abderraouf Bouziane3
Mohamed El Bachir El Ibrahimi University of Bordj Bou Arreridj, Algeria1,3, Mohamed El Bachir El Ibrahimi University, Algeria2

Keywords
Finger Knuckle Print, Major, Minor, PCAnet, Score Level Fusion, SVM
Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 10 , Issue: 3 )
Date of Publication :
February 2020
DOI :

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.