Online E-Voting system is a convenient process to poll their votes easily for users. An online voting system is a very popular method but it has a lot of security issues. The major issue of this process is considered a user authentication and data storage. There are many secured passwords based securities are available in the last decades, but such kind of securities are not safe for user authentication it is easily hacked by the attackers. Hence, Biometric based E-voting system provide reliable security for the confidential E-voting system. The retina based E-voting has provided more reliable security for user authentication. In this work, the feature extraction process is done by the fuzzy logic and the matching process is done by the hamming distance and Manhattan distance to match frequent patterns in similarity measures between the retinal images and identifies detection probabilities in retina layers. Angular and radial partitioning techniques are used to detect the similarities of the blood vessels. This mechanism provides maximum security and achieves optimal results for the E-voting system.