Publication: Estudio del rendimiento biométrico de sistemas de huella dactilar: análisis de diferentes sensores y algoritmos
Loading...
Identifiers
Publication date
2015-09-25
Defense date
2015-10-16
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Este proyecto tiene como elemento principal los sistemas de reconocimiento biométrico basados en huella dactilar. Estos sistemas, son altamente utilizados e implementados en aplicaciones de todo tipo como para el control de accesos o forenses. Por lo que en este TFG se realizará una evaluación de rendimiento de un sistema de reconocimiento biométrico analizando dos algoritmos, con tres sensores de huellas basados en distintas tecnologías: uno térmico y dos capacitivos.
La realización de la evaluación se ha efectuado a partir de una base de datos obtenida durante las prácticas en empresa cursadas en el Grupo Universitario de Tecnologías de la Identificación, en la que se obtuvieron las huellas de los dedos pulgar, índice y corazón de ambas manos de 589 individuos.
El trabajo realizado ha sido el procesamiento de la base de datos desarrollando una aplicación en el entorno Microsoft Visual Studio 2013 capaz de ejecutar la comparación de las muestras obtenidas por cada sensor con cada algoritmo.
Los resultados son almacenados en diferentes ficheros, correspondientes a cada sensor y algoritmo, realizando la distinción entre genuinos e impostores.
Con un software implementado en Matlab se realiza la segunda parte del proyecto: el procesamiento de los ficheros para obtener las medidas de rendimiento gráficas del análisis realizado.
El presente documento describe el diseño y desarrollo del trabajo realizado, junto con los resultados obtenidos tras efectuar el análisis de rendimiento.
This project’s main focus is the biometric recognition systems based on fingerprint, which all kinds of applications take advantage of, such as control access or forensic realms. Therefore, this TFG includes a performance evaluation of a biometric recognition system, made by assessing two dedicated algorithms. Three fingerprint-based sensors, which belong to different technologies has been used: one termic and two capacitive sensors. A Data base, generated during some internship within the “Grupo Universitario de Tecnologías de la Identificación”, was the driver of the assessment. The result was a collection of fingerprints (thumb, fore finger and middle finger) from 589 different people. Microsoft Visual Studio 2013 was the selected environment to process the Data Base by means of a purpose-built application that is able to compare per sensor and per algorithm every sample to each other. The genuine-false detections were stored in separated files, each one corresponding just to one sensor and one algorithm. The second part of the project conceived the use of Matlab in order to produce a turnkey software module to process those files. This process produced the graphical performance measurements of the assessment. Summarising, this document depicts the work’s design and development as well as it collects the performance assessment’s results.
This project’s main focus is the biometric recognition systems based on fingerprint, which all kinds of applications take advantage of, such as control access or forensic realms. Therefore, this TFG includes a performance evaluation of a biometric recognition system, made by assessing two dedicated algorithms. Three fingerprint-based sensors, which belong to different technologies has been used: one termic and two capacitive sensors. A Data base, generated during some internship within the “Grupo Universitario de Tecnologías de la Identificación”, was the driver of the assessment. The result was a collection of fingerprints (thumb, fore finger and middle finger) from 589 different people. Microsoft Visual Studio 2013 was the selected environment to process the Data Base by means of a purpose-built application that is able to compare per sensor and per algorithm every sample to each other. The genuine-false detections were stored in separated files, each one corresponding just to one sensor and one algorithm. The second part of the project conceived the use of Matlab in order to produce a turnkey software module to process those files. This process produced the graphical performance measurements of the assessment. Summarising, this document depicts the work’s design and development as well as it collects the performance assessment’s results.
Description
Keywords
Biometría, Huella dactilar, Sensores, Algoritmos, Medidas de rendimiento, Biometric, Fingerprint recognition, Sensors, Algorithms, Performance measures