Identificação e autenticação por análise computacional da geometria da mão.

While providing services that require access control configuration, companies are facing problems to establish an association between an individual and an identity. Recently, Web services for authentication based on politics of password, data cryptograph’s mechanism and digital signature are excessively operated. However, these methods still show vulnerabilities that result in a high rate of false verification, even in modern systems. In order to upgrade the level of information system security, the BioPass project is being developed. The BioPass predicts the utilization of multifactor authentication and multibiometrics related to active security. The goal of this project is to search and develop a subsystem of BioPass called HandPass that focuses on biometric recognition using hand geometric analysis. The HandPass excel for the huge functionality that put effect through innovative solutions and software architecture, which is open and reconfigurable. Due the lack of appropriated hand image bases, the research required, firstly, constructing a capture device, and secondly, starting to make a database for images of hands. Subsequently, all the acquired images have passed for a phase of preprocessing where they were segmented for being compared by the FullCSS algorithm. After the first tests have been processed, it was observed that the finger and wrist positions in different hand images of the same person decrease its correlation coefficient. In order to solve this problem, the wrist was removed from the images and an algorithm was developed for searching picks and valleys on fingers, and consequently, through a Warping, all the fingers could be placed in a single position improving the results of comparisons. To further projects, it is expected the increasing of images in databases with achieving more tests and adjusts that have been needed and the integration to other subsystem of BioPass focused on biometric recognition of the palm printing.

Grantor

Ciência da Computação

Year dc:date.issued

2013

Author dc:creator

• Barros, Jéssica Urbano Pereira de.

• Biometria
• Identificação biométrica
• Autenticação biométrica
• Biometry
• Biometric Identification
• Biometric Authentication

Language dc:language.iso

pt

Repository record dc:identifier.uri

https://repositorio.ufpb.br/jspui/handle/123456789/492

OAI identifier oai:identifier

oai:repositorio.ufpb.br:123456789/492