Parametric-Based Facial Recognition Technique for Improved Electronic Voting System

Abstract

Voting process plays a key role in the national development, with the attendant cost of time and other resources in the election. Recently electronic voting is gaining ground in some African countries with full benefit of accuracy and data reliability. The work proposed a biometric facial recognition strategy driven with tokenized unique parameter authentication techniques for electronic voting that reduce time, expenses, and human effort. This system is more secure as it uses the mechanism of a multifactor authentication verification process implemented using a unique token sent through SMS. Results generated from the proposed system offered very precise and improve accuracy and performance. Election reliability is enhanced, confidentiality, time and cost effectiveness are achieved. Further result gotten from the system deployment shows that 40% and 32.6% of the respondents asserted strongly and strong respectively, that the new system can perform life face detection. In addition, 56% of the stakeholders considered accepted that the new system can work effectively with unique token in less than 5 minutes.