Automatic vehicle access control into the residential areas using Face Detection and ANPR technologies

This research paper presents a system designed to enhance security and access control in residential complexes. Leveraging a suite of cutting-edge technologies and algorithms, this solution automates the vehicle access process with remarkable precision, effectively safeguarding the premises while providing user-friendly functionality. The system initiates as a vehicle approaches the entrance, employing infrared sensors to trigger the process. The vehicle’s license plate is captured and subjected to Optical Character Recognition (OCR) using K-Nearest Neighbors (KNN) and Convolutional Neural Networks (CNN) algorithms. When the license plate is cross-referenced with the database, the system proceeds to the next layer of security: facial recognition, powered by the Local Binary Pattern Histogram(LBPH) algorithm. If both the license plate and the driver’s face are successfully recognized and aligned with the authorized users’ database, the gates open seamlessly, granting access. In cases where the license plate does not match, the system promptly scrutinizes the driver’s face. Unauthorized access attempts are met with swift denial, reinforcing security. The system is evaluated by building a prototype using an HP webcam, a toy car, a controlled gate, etc. As it stands, the system boasts an impressive $96.54 \%$ accuracy for facial recognition and $88.67 \%$ accuracy for Automatic Number Plate Recognition (ANPR).