Contact-Free Multispectral Identity Verification System Using Palm Veins and Deep Neural Network.

Area of Science:

• Biometrics
• Computer Vision
• Security Systems

Background:

• Biometric systems offer convenient and secure user authentication using unique physical characteristics.
• Enhancing security often involves multi-modal biometric approaches to improve accuracy and prevent fraud.
• Palm vein recognition is a highly secure biometric modality due to the uniqueness and internal nature of vein patterns.

Purpose of the Study:

• To develop and evaluate a novel multi-modal biometric system combining palm vein imaging with IR and UV wavelengths.
• To enhance user authentication security and accuracy by integrating two distinct biometric data sources.
• To assess the system's performance in a real-world verification scenario.

Main Methods:

• Utilized infrared (IR) and ultraviolet (UV) wavelengths for capturing palm vein images.
• Employed a deep convolutional neural network (CNN) for feature extraction and user authentication.
• Conducted a verification test comparing captured images against a database of known biometric features.

Main Results:

• The combined IR and UV palm vein system demonstrated high accuracy in user authentication.
• Achieved a True Positive Rate (TPR) of 99.5% when the acceptance threshold was set at the Equal Error Rate (EER).
• The multi-modal approach significantly enhanced the reliability of the biometric system.

Conclusions:

• The proposed IR and UV palm vein biometric system offers a robust and secure method for user authentication.
• Multi-modal biometric fusion, using IR and UV wavelengths, effectively increases system security and performance.
• This technology provides a promising solution for enhancing security in various applications requiring reliable user identification.