SMART DOOR DESIGN WITH FACE AUTHENTICATION BASED ON ESP32-CAM AND WEBSOCKET

Abstract

Door access security is an essential requirement in modern housing systems. Conventional methods, such as mechanical keys and access cards, are considered less effective due to the risks of loss and duplication. This study designs a smart door system based on face authentication by utilizing ESP32-CAM for image acquisition, a servo motor as the lock actuator, and an OLED display for real-time status visualization. The WebSocket protocol is implemented to support efficient bidirectional communication between the device and the server. The system operates by capturing the user’s facial image via ESP32-CAM, performing verification with a registered database, displaying access status on the OLED, and controlling the servo accordingly. Experimental results indicate that the system achieved a 92.7% recognition accuracy with an average response time of 1.84 seconds. These findings demonstrate that the proposed design effectively enhances both security and convenience of IoT-based door access control.