Palm vein recognition based on competitive coding scheme using multi-scale local binary pattern with ant colony optimization

Yassir Aberni

doi:10.1016/j.patrec.2020.05.030

Return

Palm vein recognition based on competitive coding scheme using multi-scale local binary pattern with ant colony optimization

2020-08-01

Yassir Aberni

Larbi Boubchir *

Boubaker Daâchi

DOI：10.1016/j.patrec.2020.05.030

Original

Save

View PDF

Abstract

En 中文

Among the various biometric traits that can be extracted from the hand, the palm vein structure that represents a reliable and secure source for identifying and/or verifying the identity of a person. Several recognition methods were proposed in the literature exploiting this modality; among them, the attractive approaches based on a competitive coding. Aiming to further improve the performance of these approaches, this paper presents a novel palm vein recognition method for personal authentication and identification based on a competitive coding scheme using Multi-scale local binary pattern (MLBP) with Ant colony optimization (ACO). ACO allows to override potential blocking points related to image quality or contrast problems that can be encountered with images from the Near infrared spectral band. Thepre-processed images will be then sorted with a competitive coding scheme using MLBP; where the final image will be composed of the winning code from the different MLBP images. The matching process for making-decision is then performed using Kullback-Leibler divergence and Jaccard distance. The experimental results obtained on MS-PolyU database has shown that the proposed method achieves improved performances for both identification and verification modes up to 99.64% in terms of CIR for the identification and 0.0 0078% in terms of EER for the verification; and also outperforms the state-of-the-art methods. (C) 2020 Elsevier B.V. All rights reserved.

Keywords:

Biometrics

Palm vein recognition

Competitive coding

MLBP

Ant colony optimization

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.