Αναγνώριση ανθρώπων σε θερμογραφικές εικόνες

Βιολάτος, Νικόλαος Δ.

Master Thesis

Author

Βιολάτος, Νικόλαος Δ.

Date

2013-05-16

Abstract

The biometric characteristics of humans, such as face, speech, hands, body, fingerprints, iris color of eyes, etc., characterize unique. The most popular of these is the face and corresponding detection and recognition techniques, which are easier and more effective than other recognition techniques of other biometric features. Even the equipment is less expensive to implement a system for the recognition of people in classic photos. It is needed a simple computer and a camera. To identify on thermographic images required a thermal camera, which has a very high cost and varies the possibilities. The identity of the person has become a field of research likely associated with the recognition. Facial recognition differs from the identity of the person because the first has as a "duty" to determine the identity of an object, while the latter need to verify the identity of a user. In this work we give weight to recognize human on facial images, i.e. face recognition. The images used will not be the classic color, black and white or grayscale ones, but thermographic. The advantages of thermographic images that are not affected by ambient light conditions and are not altered properties and attributes if the person being photographed is wearing a hat, glasses or masquerading. In thermographic images look temperatures unlike conventional photographs illustrating color information with brightness values. The algorithms and techniques used were selected in such a way that they can be applied to conventional images and the proposed system to be as scalable as possible.