Biometric Authentication in E-Voting through RNN

Abstract

In a democratic country the election process and the right to vote are one of the most significant aspects. This is due to the fact that in a democratic country the sole administrator and the decision maker for the entire country needs to be elected in a fair and just manner from amongst the citizens of the country. This procedure has been effectively performed manually and physically by the utilization of ballot elections. This is a highly inefficient form of election that needs to be upgraded in this day and age of information and electronic supremacy. But the considerable concerns for developing an electronic voting scheme was the security concerns regarding multiple voting performed by a single person. Therefore to introduce an effective and useful technique which also addresses the security concerns an effective methodology for electronic voting through facial recognition has been proposed in this research article. The proposed methodology utilizes the open-source open CV library along with Recurrent Neural Networks to achieve highly accurate facial recognition for a secure electronic voting system. This approach has been significantly e evaluated for their performance metric which has achieved suitable results.