Design and implementation of a multibiometric system based on hand's traits

In this paper, we present a multibiometric system based on two different biometric modalities, palmprint and fingerprint. The proposed system is based on a score fusion which can effectively combine the discriminative power of multiple biometric traits. We propose a combined fingerprint image enhancement, which achieves satisfactory results. We also apply two different indexing algorithms on fingerprints and three different indexing algorithms on palmprints. Existing multibiometric systems are not sufficiently fast for practical applications. Indexing and retrieval mechanisms can effectively narrow down the search space in large databases. We propose a fingerprint indexing based on ellipse properties, which reduces false correspondences. The proposal uses a k-means clustering algorithm to index and retrieve fingerprints and palmprints. Furthermore, we combine some minutia cylinder-code (MCC)-based algorithms and improve the MCC-based indexing algorithms. Empirical experiments on some of datasets of the national institute of standards and technology (NIST), fingerprint verification competition (FVC), and THUPALMLAB databases show that the proposed algorithm considerably narrows down the search space in fingerprint and palmprint databases and increase the recognition performance.