The electronic voting system is the easiest way for the election process. The E-Voting System is a digital electronic system. In this process, the user’s data are collected in a digital manner. The stored data should be protected from unauthorized persons. The major issues of this process are the protection of stored data. The data security is considered as an important factor in an online voting system. Hence, the stored templates should be needed to prevent unauthorized users. In this work, explores unique retina features in a single can be existing in a binary format which can be quickly matched with the stored template to authorize identity. As biometric template are deposited in the database, owing to security threats biometric template may be altered by unauthorized persons. If the biometric template is altered authorized user will not be allowed to access the source. A box counting algorithms used to extract retina blood vessels from retina image. It analyzes and presents a high-level security and better image encryption using MD5 algorithm which is implemented on high descriptor value in an image. The matching process is done by using a tree data structure. The experimental results show high accuracy for the matching process when applying more images and achieve optimal results in secure online E-voting system.