Embedded fingerprint verification system

• Main Author: Zhu, Xu.
• Other Authors: Wen, Changyun
• Format: Theses and Dissertations
• Published: 2008
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Biometrics
• Online Access: http://hdl.handle.net/10356/4088
• Institution: Nanyang Technological University

The objective of this thesis is to develop an embedded fingerprint verification system using an e-field fingerprint sensor and DSP processor. The e-field fingerprint sensor is an advanced TRUE fingerprint image sensor from AuthenTec. INC. The aim of this development is to speed up the fingerprint verification algorithms and adapt those algorithms to DSP platform. This thesis investigates and simulates the performances of fingerprint verification algorithms including pre-processing, enhancement, minutia extraction, and minutia matching, and improves the performance of point matching algorithms and image pre-processing based on the e-field fingerprint sensor's properties. An embedded fingerprint verification system has been proposed with hardware and software design considerations. The design and implementation of these techniques are documented. Some recommendations based on the experience gained from the project are discussed.