Paper Title
An Embedded Platform For Automatic Teller Machines Using Finger-Vein Recognition System

In this paper an embedded finger-vein recognition system for authentication in Automatic Teller Machines has been proposed. The personal ID (Finger vein image) of the user is initially inserted. The verification of the finger vein image is done using a finger vein image matching technique. The finger vein image of the user is compared with the authenticated finger vein images of the data base. The finger vein image matching of the ATM user is done after a series of processes: resizing, image segmentation, histogram equalization and feature extraction. The process of feature extraction is done by taking five features: mean value, singular value decomposition, energy, entropy and variance. An instant PIN is generated, which is sent to the mobile of the authenticated user to provide additional security for the transaction.