An overview of uni- and multi-biometric identification of identical twins

Abstract

Biometric techniques can differentiate identical twins, and this makes the techniques a subject of interest. Studies on biometric techniques have been conducted in many disciplines. Unimodal biometric systems with twins, however, have issues including intra-class variations, noisy data, limited freedom levels, non-universality, vulnerability to spoof attacks, and intolerable rates of error. Several of these limitations are manageable through the use of multimodal biometric systems that combine the evidence brought by many sources of information. This survey highlights the relevant literature, the accessible datasets for research, and the subjects of uncertainty, and proposes future research.