The current electronic voting system can be hacked easily. There are a lot of
methods adopted to avoid malpractice. This research provides secured voting and
avoids human intervention that results in smooth and secure conduction of elections.
This research adopts biometric fingerprint recognition and face recognition of the voter
for authentication. In an electronic voting system, the first step in the verification
process can be easily achieved with the voter fingerprint data available in this database.
The second step of verification involves the face recognition of the voter by the data
already present in the database. If two-phase verification is done, the voter can proceed
with the voting process and present his/her vote. Then the vote will be encrypted. This
prevents fake votes and ensures perfect polling without any corruption. We have
created a fingerprint-based voting system where the user does not have to take hisher
ID with his/her necessary information. If the details match the previously stored
information of registered fingerprints, a person is allowed to cast his or her vote. If not,
a warning message will be displayed and the person is excluded from voting. In an
election counting stage, the admin will decrypt and count the votes.
Keywords: Deep learning algorithm, E-Voting, Encryption and decryption, Fingerprint recognition, Facial recognition.
