Study of the intra-variability of a person’s face from infrared images

This is a report for the Final Year Project held in the final year of study of the u-year Electrical and Electronics Engineering course in Nanyang Technological University (NTU), starting in August 2009. Machine recognition of human faces from still images and video images has increased their re...

Full description

Saved in:
Bibliographic Details
Main Author: Liao, Marc Zhi Xiong.
Other Authors: Sung, Eric
Format: Final Year Project
Language:English
Published: 2010
Subjects:
Online Access:http://hdl.handle.net/10356/40492
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:This is a report for the Final Year Project held in the final year of study of the u-year Electrical and Electronics Engineering course in Nanyang Technological University (NTU), starting in August 2009. Machine recognition of human faces from still images and video images has increased their research activities extensively over the past 30 years. The face has been a main factor in researching area since traditional means of security such as Identification (ID) cards. Face recognition offers a non-intrusive but maybe the most natural way of identification. While human can recognize faces easily despite several years of separation, the problem of machine face recognition is still a challenging task. Therefore the potential for illumination invariant face recognition using infrared (IR) images has began. Many researchers have attempted to use IR images for face recognition but it is not clear that such a method will ever work. There are no study of the intra-variability of the same person face has been done. This study is to show the possibility of intra-variability and inter-variability to be successful so that this objective can be achieved. A program for intra-variability and inter-variability is designed to test whether this analysis would be successful. This program is analyzed with a set of IR images. These IR images are taken with five people for ten times at different time and day throughout the project. Recognition is based on theory of Zero-Centered Normalized Correlation Coefficient (ZNCC). This recognition is based on template matching method as this is one of the easiest methods. As I want to achieve my objective for this report, this study would be a small step to test whether the intra-variability of the same person is successful. Further analysis will be explain and discuss in detail, together with the possible problem and solution to it.