Twins multimodal biometric identification system with aspect united moment invariant

Abstract

In the field of pattern recognition twin’s biometric identification is currently a popularly studied subject. In some situations, the mechanism of twins’ biometric Identification leads to the finding a distinctive pattern of a person’s biometric. Correspondingly, there has been considerable improvement made on the Unimodal biometric identification to identify identical twins with respect to its accuracy and reliability, with some traits that show sound performance. However, owing to great level of similarity, it is much more challenging to identify Identical twins as opposed to identifying non-twins. In order to deal with this problem, the application of more than one biometric trait is proposed; the Multimodal biometric system. Meanwhile, in pattern recognition it is crucial to extract and select features that are meaningful. This brings the attention to the major issue in twin handwriting-fingerprint identification: how to obtain features from numerous writing and styles twin handwriting-fingerprint so that the right person between twins can be reflected. Hence, the Aspect United Moment Invariant is proposed in this study as extraction of feature with identical twin multi-biometric identification.