<![CDATA[This study aims to design and implement an Internet of Things (IoT)-based server room access logging system using facial recognition technology. Conventional security systems such as mechanical locks and access cards have several limitations, including vulnerability to loss, duplication, and unauthorized access, as well as the absence of automatic activity logging. Therefore, this research proposes an integrated system that combines facial recognition for authentication with real-time access logging and monitoring. The system utilizes ESP32-CAM as an image acquisition device, while face detection is performed using the Haar Cascade method and identification is conducted using cosine similarity. Data collection was carried out by capturing 50 facial images per registered user at a distance of approximately 30 cm with variations in facial angles. The system was tested under real operational conditions, showing an average accuracy of 91% with a response time of approximately 1–2 seconds. In addition, the system successfully records access activities, including access status and user data, which can be monitored through a web-based interface. The results indicate that the proposed system can improve server room security by providing automated authentication, access control, and real-time monitoring in an integrated manner.]]>
Copyrights © 2026
