A Privacy Preserving Based Multi-Biometric System for Secure Identification in Cloud Environment

Privacy-preserving biometrics identification has drawn greater attention that efficiently provides secure storage of sensitive information in cloud servers. In this paper, we design a Privacy-Preserving Multi-Biometric Identification (PPMBI) scheme that ensures secure biometric data outsourcing and protects user privacy against higher-level attacks in the cloud. This approach uses a multi-biometric model that integrates the fingerprint and finger vein features to improve data security. Also, we present a weighted sample encryption process and the user profile encryption that controls access of user data and finds a close match for the input query. During the user identification stage, pattern matching and decryption with the minimum distance computation are performed to obtain the index of the user registered in the database. Based on the computed index value, similarity is estimated and outputs the result as identified or denied to the user. Experimental analysis indicates the proposed approach achieves better privacy and resists all possible attacks in the cloud system.