Abstract:Recently, finger vein recognition has become integral to biometric technology that uses vein patterns for highly secure personal recognition. Many existing studies reported accurate finger vein recognition systems using recent techniques. However, the existing techniques face high complexity in analyzing vein patterns and increase recognition errors. Hence, this article proposes a novel deep learning (DL) technique for automatically recognizing finger veins at low time complexity. Initially, image de-noising and contrast enhancement are performed in the preprocessing stage using Extended Frost Filtering (Ex-FF) and Entropy Enriched Histogram Equalization (En-HE) to enhance the image's visual quality. The preprocessed images are then fed into the Kernel Attentive Resnet152v2 (KAtt-ResNet152V2) technique to extract the vein features efficiently. The extracted vein features are then fed to the Hybridized Golden Archerfish Optimizer (HyG-ArfO) to select the optimal vein features to reduce the feature dimensionalities. Finally, the optimal features are fed into the Residual Pyramid based Depthwise Separable MobileNetV3 (PyDS-MV3) technique to accurately recognize the person's finger veins. The proposed work is implemented in MATLAB, and a publicly available finger vein dataset is utilized. The performance measures such as accuracy, precision, false discovery rate (FDR), and Mathew's correlation coefficient (MCC) are analyzed and compared with existing studies. The proposed method obtains an overall accuracy of 99.6%, precision of 99.42%, FDR of 0.005 and MCC of 0.994.

New finger biometric method using near infrared imaging

A Novel Finger Vein Pattern Extraction Approach for Near-Infrared Image

Multimodal Biometric Identification System Based on Finger Geometry, Knuckle Print and Palm Print

The Biometric Recognition System Based on Near-Infrared Finger Vein Image

A New Biometric System Based on Inner-Knuckle-print Recognition

Survey on Multimodal Biometric Authentication Combining Fingerprint and Finger Vein

Near Infrared Finger Vein Recognition Method Based on Subspace Projection Technology

Personal Identification for Single Sample Using Finger Vein Location and Direction Coding

A New Finger Vein Recognition Method Based On Lbp And 2dpca

Deep Belief Network Based Finger Vein Recognition Using Histograms Of Uniform Local Binary Patterns Of Curvature Gray Images

Robust Hybrid Finger Pattern Identification Using Intersection Enhanced Gabor Based Direction Coding

Significance of Being Unique from Finger Patterns: Exploring Hybrid Near-infrared Finger Vein and Finger Dorsal Patterns in Verifying Human Identities.

A Method for Capturing the Finger-Vein Image Using Nonuniform Intensity Infrared Light

Personal identification based on multiple keypoint sets of dorsal hand vein images

From Noise to Feature: Exploiting Intensity Distribution as a Novel Soft Biometric Trait for Finger Vein Recognition

A Finger-based Recognition Method with Insensitivity to Pose Invariance

Small-Area Finger Vein Recognition

Fusion of Band Limited Phase Only Correlation and Width Centroid Contour Distance for finger based biometrics

Personal authentication using finger vein pattern and finger-dorsa texture fusion

An optimized deep learning based depthwise separable MobileNetV3 approach for automatic finger vein recognition system

A Finger Vein Authentication System Based on Pyramid Histograms and Binary Pattern of Phase Congruency