| Popis: |
Electronic voting systems have often associated with issues of security, transparency, and limited accessibility, particularly for people with disability, elderly, and visually impaired voters. These systems, impeded by complex authentication procedures and non-intuitive user interfaces, have also struggled to ensure voter trust and widespread participation. To address these issues, this paper presents a new electronic voting system that combines the gesture recognition with blockchain technology. The proposed system has two phases, the first one is a user identification phase using a pre-trained dataset based on both cvzone and Dlib tools for face and finger identifications. Gestures from the face and hands, such as eyebrow and nose movements are used as authentication mechanisms as well as a mean of input into the system. The second is a voting process phase which uses the blockchain technology. This new approach strengthens the security and transparency of voting through a tamper-proof ledger provided by the blockchain technology, as well as enhances considerably the inclusiveness and also user friendliness. Real time simulations show that the user interaction times for different age groups ranges between (24.3s to 29.7s) and the accuracy rate for gesture types varies between (88.7s to 92.5s). Moreover, the simulation results indicated that real time voting using our proposed system generated 100% precision and 99% recall. Our system showed high accuracy in gesture recognition and profound security when using Equal Error Rate quality measure. Different to other approaches, the interface of our proposed system, designed for a wide audience of users, greatly simplifies the voting process. |
