A multibiometric finger vein verification system based on score level fusion strategy

Multibiometric systems are very applicable in human identification systems nowadays. However, using different biometrics may need extra acquisition sensors besides algorithm complexity. In this paper, a new multi-instance human identification system is proposed using different finger vein biometrics. In the proposed method, the multibiometric system is implemented using score level fusion. A complete investigation is done on effectiveness of different score level fusion strategies. For the proposed multi finger vein biometric system experimental results confirm that maximum score level fusion strategy outperforms other fusion methods in addition to some classical fusion algorithms in terms of EER.