Studies in pattern recognition domain currently revolve around twins biometric identification. The twins biometric Identification system may lead to the discovery of a distinguishing pattern of a biometric of an individual. A significant improvement can also be seen in the Unimodal biometric identification, it allows accurate and reliable identification of identical twins with good performance of certain traits. However, since the similarity level is very high, Identical twins identification is much more difficult when compared to that of non-twins. Hence, the use of more than one biometric trait with global features is proposed. Further, pattern recognition requires the extraction and selection of meaningful features, which leads to the key issue in the identification of twin handwriting-fingerprint, that is, the question of how to acquire features from many writing and styles twin handwriting-fingerprint to enable the reflection of the right person between twins. This study thus proposes the global with Aspect United Moment Invariant for global feature extractions with the application of identical twin multi-biometric identification with Inter-class and Intra-class .
展开▼
