Real Time Finger-Vein Recognition For Automatic Door Access Control

Ngan, Albert Ban Chew (2017) Real Time Finger-Vein Recognition For Automatic Door Access Control. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Real Time Finger-Vein Recognition For Automatic Door Access Control.pdf - Submitted Version Download (488kB)

Abstract

This project is to build a real-time finger-vein identification device for door access control.It is a device to capture the finger-vein image of a person to give access to the authorized user.A matrix of near-infrared light emitting diode (NIR LED) will be emitted on to the finger in a box whereby a CMOS NOIR Raspberry Pi camera on the other side of the finger will be ready to capture the finger-vein image.The veins would appear darker in the image whereas the background or the finger will be lighter.The image will then go through a series of image processing steps to enhance the image and lastly taken it’s minutiae points to compare with the template in the database using Modified Hausdorff Distance (MHD). If the user is authorized, door will open and vice versa. The device is made by using a Raspberry Pi 3 as the microcontroller to process the image,controlling the camera,NIR LED’s lighting and to control the motor of the door.The Equal Error Rate (EER) of the device is 7.88% which yields an accuracy of 92.12%.

Actions (login required)

View Item