Palm Vein Recognition Using Networking

Abstract

Palm vein identification relies on unique patterns within the palm veins, illuminated by Near Infrared Light (NIR) with wavelengths from 760 nm to 820 nm, penetrating the skin up to 5mm. Absorption of NIR by deoxygenated blood in veins creates distinct dark patterns. However, this high wavelength light may cause skin and tissue infection. Vein networks are captured via infrared-sensitive cameras, with captured images pre-processed to remove noise and features extracted for recognition. Feature extraction primarily involves network segmentation, creating reference maps for subsequent recognition. These feature maps serve as blueprints for neural networks, facilitating streamlined identification processes.