Fusion of Speech and Iris in Multibiometrics

摘要

Biometrics is constantly evolving technology which has been widely used in many identification applications. Biometric modalities are divided into two parts as unimodal and multimodal. In the real world applications, unimodal biometric modalities have many problems and under highly unconstrained conditions, it is not possible to get desirable results. Thus the researchers have moved towards the development of multimodal biometric system. In this regard, multibiometrics is used in high security systems. Fusion in multibiometrics is required to improve the performance of biometric system. Multibiometric is a challenging problem as it is very difficult to predict the optimal fusion strategy due to correlation and fusion complexity. In many multibiometric systems, the different biometric sources (speech and lip movement, two impressions of a person's fingers) are correlated. In this paper, the limitations of unimodal biometric modalities and the types of multimodal biometrics are discussed. In this work, the fusion strategies in multibiometrics are analyzed with the implementation of speech and iris.