Multibiometric Secured E-Voting Based on Hyper Elliptic Curve Cryptography

Abstract

In popular democratic countries, the voting system plays an important part during choosing the right government. Electronic voting machines are less secure with regard to data, time consuming and need further workforce to avoid illegal casting of votes. The proposed system have a secured authentication method with a multibiometric system integrating both fingerprint and iris recognition system. The stored iris data is encrypted using hyper elliptic curve cryptography (HECC) algorithm. The Daugman’s algorithm smoothen the noise posterior to the iris localization and feature extraction is carried out to non-varying and unique parameters of an iris image. Further biometric data used is fingerprint recognition, which is a complex recognition parameter. It includes algorithms and procedures for image improvement and linearization, rooting for miniaturized features and matching for authenticating the candidates. To ensure security, iris images converted to cipher templates using HECC, thereby data can be saved from duplication without the chance of intrusion. The proposed system witnessed a true acceptance rate of 98% with 2.5% of equal error rate and accuracy of 95%. The minimum recognition time achieved is 4 seconds and has been identified as an outperforming model than existing solutions in voting applications.