Implementação de verificações biométricas processadas em cartões inteligentes a multi-aplicações.

• Autor(a) principal: Wyant, Rafael Soares
• Data de Publicação: 2013
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UERJ
• Texto Completo: http://www.bdtd.uerj.br/handle/1/11828
• topic: Electronic Engineering; Biometrics; Smart cards; Minutiae; PalmCode; IrisCode; Engenharia Eletrônica; Biometria; Cartões inteligentes; Minúcias; PalmCode; IrisCode; CNPQ::ENGENHARIAS

The biometrics have been used as a solution for access control systems for many years, but the simple use of biometrics can not be considered as final and perfect solution. There are many risks that should not be ignored. Most problems are related to the transmission path between the system where the users require access and the servers where the captured biometric data is stored. Various types of attacks can be made by impostors who want to use the system improperly. Besides the technical aspects, there is the social aspect. There is a growing concern of users about both data storage and the misuse of their biometrics, which is an unique identifier and, being invariant in time, may be lost forever if compromised. The fact that several companies keep their biometric data in different servers is causing discomfort to users because it makes their biometric data more susceptible to attacks. In this dissertation, the use of smart cards is adopted as a possible solution to the above problems. Smart cards prepared for multi-applications are used to perform biometric comparisons internally. Thus, it would not be necessary to use different servers as the biometric features will always be kept on a single card in the possession of the owner. In this work, three different algorithms using different biometric identification characteristics are developed and implemented: fingerprint, palmprint and iris. Considering the used memory, average execution time and accuracy, the implementation of palmprint verification achieved the best results, allowing minimum error rates and processing time of at most 0.5s.