Análise odonto facial para identificação biométrica

Abstract

The process of human identification or biometric identification consists of differentiating one person from another through its unique characteristics (physical, psychic, functional and civil). The most cited biometric techniques are: fingerprint, retina, pulse and facial recognition. Normally, the areas of the face detected in the recognition process are the eye, nose, mouth and chin. This paper addresses the first phase of facial recognition - facial detection, and proposes to identify, in a digital image, the upper canine tooth´s characteristics in: normal (N), long (L), short (S) and inclined in its long axis (I) and to detect the facial structures eye, pupil, nose, mouth, isolating them in individual cuts. Free tools such as images of open database Label Face in the Wild (LFW) from the University of Massachusetts, MatLab ((MATrix LABoratory) software, Paint and microsoft office excel 2007 have been used in the development of this work. As a result, a higher prevalence of the normal characteristic was observed in relation to the others, as we also observed that the same image may present more than one characteristic. The importance of this work is to offer one more tool to the healthcare professional, capable of assessing and identifying position anomalies, long axis alterations and upper canine teeth shape alterations.