Optimal Double Layer Secret Sharing Scheme for Biometrics

Abstract

Visual secret sharing (VSS) method is an encryption method to ensure the Security of secret data, that just performs the partitioning of ‘n’ shares and distributes among ‘n’ users in an ideal way so that exposing of less than ‘n’ shares is of no utilization. However, some dishonest members or hackers team up and tries to cheat other members in the group. Therefore, we have developed a dual layer secret sharing scheme based on universal share based secret sharing scheme. The dual layer is composed of threshold based secret sharing and then followed by the universal share based secret sharing. Here, the universal share is maintained by the trusted party avoids the contribution of dishonest participants. Moreover, to ensure additional security, the proposed approach employs Oppositional Artificial Fish Swarm Optimization (OAFSO) based Stream Cipher encryption technique to encrypt the shares, shows the novelty of the work. Furthermore, the confidentiality is enhanced with Biometric fingerprint authentication step, where the acknowledged users are alone allowed to get the decrypt shares with that the user can retrieve the secret data. The proposed fingerprint authentication method also makes use of Secure Hash Algorithm (SHA 1) to store the fingerprints and the matching is also done with hashed fingerprints only. So that the attackers cannot take and corrupt the stored fingerprints and the misuse of fingerprints is not possible. Finally the performance analysis is made with existing approaches in terms of PSNR and MSE. Maximum PSNR is 58.9802 and minimum MSE value is 0.6232, while the existing methods provide very less PSNR and greater MSE values than the proposed methods.