A Review on fusion in Multimodal Biometric Spoofing Attack by Different Materials

Abstract

Biometric gadgets utilize their physiological or behavioural properties for the confirmation and acknowledgment of people. Spoofing attack can be done by using any spoofing materials. Such features can be arranged into unimodal and multimodal frameworks. Some state-of-the-methods have some drawbacks, which reduce the efficiency of the system. Multimodal biometric detecting frameworks utilize at least two behavioural or physiological attributes. The multimodal system has showed to increase the success rate of identification and authentication meaningfully. Data from different modalities are acquired, pre-handled, removed noise, and contrasted and finally converted into features. At last, selection of features acknowledges the identification of a person. In multimodal biometric identification system, biometric features can club at any of the stages. i.e. sensor level, feature level, score level, rank level, and decision level. This paper presents an effective survey on fusion of features at different level in a multimodal biometric system. It also focuses in the field with a better thoughtful of multimodal biometric sensing and handling systems and research inclinations in this field.