On chip design and verification methodology for multimodal biometric applications

Abstract

Authenticating the identity of an individual has become an important aspect of many orga-nizations. This need of providing automated authentication has lead to concerns in the security and robustness of such Biometric systems. Currently, many biometric systems that are deployed in organizations are unimodal, which means that they use a single physical trait to perform au-thentication. But, these unimodal systems suffer from many drawbacks such as: noise present in the test information like scar of a body part, poor data capture because the user does not interact properly with the sensor device, presence of false matches in the authentication system which reduces the effectiveness of the system and presence of spoof attacks wherein some ma-licious user might be successful in hacking the biometric device. newline newlineThese drawbacks can be overcome by designing multi-modal systems which use multiple physical traits to perform authentication. They increase reliability and robustness of the sys-tems. The multi-modal systems are able to deliver the expected performance that is expected by many organizations. It helps in adding more number of users in the biometric process thus, increasing universality of these biometric systems. They are able to avoid spooand#64257;ng attacks by providing higher levels of security than unimodal systems. It is also considerably avoid the problem of false matching. newline newlineIn this work, a novel hardware design of multi-modal biometric systems is proposed. This system is implemented on the relevant chip hardware. It uses three physical traits namely: iris, and#64257;nger and palm. A fusion methodology has been presented which generates a single template by using all the three physical trait images. This template is generated for every database image and a user test image. newline