Presentation of Master's Thesis

Verificación biométrica a partir del acelerómetro del Smartphone

Images

Imagen 1 smallImagen 2 smallDate: September 2014
Author: Daniel Maté Arce
Tutors: Jesús B. Alonso Hernández, David Sánchez Rodríguez and Carlos M. Travieso González

 

Abstract

The recognition and identification of people has been a goal to accomplish by humans throughout history. Many biometrics and its applications have been studied.

This paper aims to deepen the study of gait as biometric modality. The presence of accelerometers in mobile devices today, makes this biometric acquire a new relevance. Whether to determine the ownership of the phone, or to detect anomalies in the gait of the device owner.

For the study of the biometric feature, creating a mobile application has been necessary. This application records the users gait, forming a database.

Likewise, a feature extraction system has been developed. This system decompose the gait signal into steps, and then extract same statistical parameters of those steps. Finally, a classic K-NN classifer has been used to predict the correct users.

When testing the system, success rates of 80% were obtained in the best case. A fact which, while not extraordinary, if allowed to be optimistic about the use of this new biometric modality.