Multimodal cumulative class-specific linear discriminant regression for cloud security

Cloud data access is one of the most active researches in all the public, private and government sectors in recent times. Authenticating cloud data is the way of securing the information present inside the cloud. Many modes of authentication are being used by the consumers for transfer and access of information, of which biometric authentication is an efficient method used for eliminating many fraudulent attacks and spoofing. This paper focus on developing a multimodal biometric authentication system, in order to improve the recognition process to a great extent without any errors. It is developed in such a way that a four way cloud generator algorithm is used to provide a strong boundary of authentication, which allows the user to access permission based on the inbuilt databases found in the system. Finally, the performance of the proposed method is also verified which shows that the proposed system performs well than the other.