What can the structure of the palmprint tell us?

With the development of technology—a double edge sword, while people are enjoying convenience and lifestyle reform, malicious usage of technology also lead to security concerns. Many identification and verification methods have been developed to address this issue. Token based solutions like pass...

Full description

Saved in:
Bibliographic Details
Main Author: Liu, Xue
Other Authors: Li Fang
Format: Final Year Project
Language:English
Published: 2016
Subjects:
Online Access:http://hdl.handle.net/10356/66613
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:With the development of technology—a double edge sword, while people are enjoying convenience and lifestyle reform, malicious usage of technology also lead to security concerns. Many identification and verification methods have been developed to address this issue. Token based solutions like password and smart card can be easily duplicated, forgotten, lost or stolen. Biometric technology is a new solution to overcome shortcomings of traditional identification methods. Biometric technology based verification and authentication regards obtaining and utilising biometric information of people (palmprint, vein pattern, facial expression) to verify and authenticate users. Palmprint and vein pattern verification are important components of biometric authentication due to their uniqueness and relative invariance along people’s life. Palmprint and vein pattern identification and verification solutions are getting more and more popular in real world application. However, there is a trade-off between identification power and efficiency which has been a bottleneck for palmprint and vein matching system. Bruteforce matching is computationally expensive. To improve efficiency, coarse-matching must be done. Here, Fourier Transform is the instrumental. However, traditional Fourier Transform usually achieve significant improvement on computational time but incur some information loss and low Matching Accuracy. Hence advanced Fourier Transform algorithms should be explored to further reduce computational time and improve Matching Accuracy. The original system contains a brute-force implementation of Discrete Fourier Transform algorithm which is inefficient and causes significant drop in Matching Accuracy. This work is concerned with developing and integrating a Mixed-radix and Radix-2 Fast Fourier Transform algorithm to reduce Transform Time and explore methods to improve Matching Accuracy.