Abstract: With the introduction of electronic voting systems, the efficiency, transparency, and accuracy of the voting process can be significantly improved. However, the conventional Electronic Voting Machines (EVMs) involve a manual verification of the voter's identity. This makes the system highly susceptible to voter impersonation and duplicate voting. This paper presents a real-time face recognition-based electronic voting system using a Raspberry Pi to improve the accuracy of the voter verification process. In this paper, the authors have designed a face recognition-based electronic voting system that includes a raspberry pi camera to acquire the voter's face images, a face detection system that uses the Haar Cascade Classifier algorithm, and a face recognition system that uses the Local Binary Pattern Histogram (LBPH) algorithm. After successful verification of the voter's identity, the voter will be allowed to cast their vote through a touchscreen interface. The system was able to attain a recognition accuracy of up to 92% under bright lighting conditions, as well as 90% under normal indoor lighting, with an average authentication time of approximately 1.3 seconds. Functional testing was equally successful, with the system being able to prevent duplicate voting attempts by 100%, as well as store the votes in the database. This shows that the proposed system is effective for use in institutional voting scenarios

Keyword: Electronic Voting Machines, Local Binary Pattern Histogram, Face Recognition, Raspberry Pi

Downloads: | DOI: 10.17148/IJIREEICE.2026.14531