Master Thesis - Department of Computer Science
Master Thesis - Department of Computer Science
Master Thesis - Department of Computer Science
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
CHAPTER 2<br />
Literature Review<br />
This research work concentrates on face recognition problem as a part <strong>of</strong> unimodal<br />
biometric system and then moves towards the combination <strong>of</strong> face and fingerprint fea-<br />
tures to obtain a robust multimodal biometric system. In this chapter, we first present<br />
some <strong>of</strong> the previous work done on face recognition in Section 2.1. Section 2.2, pro-<br />
vides a brief literature review on fingerprint recognition, and finally multiple classifier<br />
combination (MCC) techniques are reviewed in Section 2.3.<br />
2.1 Recent Approaches to Face Recognition<br />
Face recognition has been an active research area over last 30 years. This research<br />
spans several disciplines such as image processing, pattern recognition, computer vi-<br />
sion, and neural networks. It has been studied by scientists from different areas <strong>of</strong><br />
psychophysical sciences and those from different areas <strong>of</strong> computer sciences. Psychol-<br />
ogists and neuroscientists mainly deal with the human perception part <strong>of</strong> the topic,<br />
whereas engineers studying on machine recognition <strong>of</strong> human faces deal with the com-<br />
putational aspects <strong>of</strong> face recognition. Face recognition has applications mainly in<br />
the fields <strong>of</strong> biometrics, access control, law enforcement, and security and surveillance<br />
systems.<br />
The problem <strong>of</strong> face recognition can be stated as follows: Given still images or<br />
video <strong>of</strong> a scene, identifying one or more persons in the scene by using a stored<br />
database <strong>of</strong> faces [23]. The problem is mainly a classification problem. Training<br />
the face recognition system with images from the known individuals and classifying<br />
the newly coming test images into one <strong>of</strong> the classes is the main aspect <strong>of</strong> the face<br />
recognition systems.<br />
This problem seems to be easily solved by humans where limited memory can be