Biometric Data Fusion Strategy for Improved Identity Recognition

摘要

In modern authentication systems, various types of biometric measurements are used to authorize access to protected resources. Constant development of systems using biometric authentication means that they are exposed to hacker methods of stealing users' digital identities. The paper presents a hybrid system for identity identification, which uses various methods of biometric data fusion. Using the MegaMatcher environment and a set of dedicated scanners, an application was designed to allow testing of the authorization process and the impact of biometric data fusion on system security. In the experimental part of the work, two data aggregation strategies were compared, including the False Acceptance Rate (FAR) and False Rejection Rate (FRR) coefficients. The presented methods of biometric data fusion can be applied in authorization systems which use hybrid identity identification.