Secure Biometric E-Voting System Using ML

Abstract

In today's digital era, ensuring secure and tamper-proof electoral processes has become a growing necessity. This project, titled Secure Biometric E-Voting System Using Machine Learning, aims to build a reliable and user-friendly voting mechanism that verifies a voter's identity using two critical biometric iris and fingerprint alongside their Aadhar number, name, and age. By integrating machine learning into this biometric verification process, the system not only enhances authentication accuracy but also detects fraudulent or mismatched votes

To facilitate this, a custom dataset generation module was developed. It simulates a real- world biometric database by assigning each synthetic voter a unique Aadhar number, a randomly generated name, and an age within valid voting range. For each voter, fingerprint and iris images are carefully paired and organized into dedicated folders. A master summary CSV is also generated to maintain a centralized reference of all user data, allowing the system to index and cross-check voter identities during the machine learning validation phase. This dataset serves as the foundation for training a machine learning model—specifically a Random Forest Classifier to classify votes as either valid or faulty based on biometric match accuracy. The project strengthens the voting process by ensuring that only authenticated individuals can cast a vote, effectively minimizing the chances of identity theft, duplicate voting, and vote rigging.

By combining artificial intelligence with biometric data, this system presents a forward- thinking approach to electoral security, ensuring that future elections are not only smart but also safe, inclusive, and tamper- resistant. The rapid advancements in digital technology have necessitated the development of secure and reliable electronic voting systems. Traditional voting mechanisms are often susceptible to fraud, identity theft, and manual errors. This project proposes a secure voting machine that integrates machine learning algorithms to authenticate voters using biometric data—specifically, iris and fingerprint recognition and their Aadhar number.A Random Forest algorithm is employed to accurately classify and validate legitimate voters while detecting anomalies or fraudulent entries.This approach provides a robust solution to safeguard the integrity of the voting process and eliminate the risk of impersonation or duplicate voting.