Smart Vehicle Access Control Using Computer Vision-Based Face Authentication

Abstract

Vehicle security remains a paramount concern, with traditional key-based systems being susceptible to theft and unauthorized access. This paper presents the design and implementation of a prototype for a secure, biometric-based car ignition system using facial recognition. The system uses a computer vision algorithm called Local Binary Patterns Histograms (LBPH) to recognize faces. The main process captures a driver's face through a webcam. It checks this against a database of authorized users and then sends a control signal to a built-in subsystem. The ignition control hardware is built around an ESP32 microcontroller. It uses a relay to switch a DC motor, which simulates the vehicle's engine starter. Communication between the computer that performs the recognition and the ESP32 happens through a standard USB port. This framework shows a low-cost, practical option that improves security compared to traditional car ignition systems.