TOUCHLESS FINGERPRINT RECOGNITION

• Main Author: Velda, Vania
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/55939
• Institution: Institut Teknologi Bandung

Fingerprint is one of human characteristics that commonly used to identify a person in biometric authentication. Touchless fingerprint recognition represents a rapidly growing field of research as technology development. A touchless fingerprint recognition system is devised using a camera sensor to capture image. This system raises new problems such as the decreased resolution when capturing image from far and how to process the captured image so that it can be used in common fingerprint extractor and matching. Technique that can be used to increased resolution without losing details is super resolution. Captured image need to be segmented to locate the fingertip and being preprocess resulting in binary image. Minutiae extraction will be used in matching. Minutiae types ridge ending and bifurcation will be used in minutia based matching. Threshold of 40 will be set in determine whether a probe and a candidate is match. Fingerprint image quality and the number of minutiae are evaluated using FpMV software developed by NIST. A good quality fingerprint has minimum 40 minutiae and fingerprint with and without super resolution has around more than 40 minutiae with different minutiae quality. Super resolution does not increase the biometric performance as the CER is fairly high compared to system without super resolution. With super resolution, CER value is around 30%-40%.