Dual Sistem Keamanan Pada Pintu Dengan Pengenalan Wajah Local Binary Pattern Histogram (LBPH) Dan Sidik Jari serta Notifikasi Telegram

<p><strong>Conventional door security systems, such as padlocks and manual keys, have weaknesses, including vulnerability to duplication and the risk of loss. Biometric-based systems, such as facial recognition, offer a more reliable solution through unique user identification. This study develops a door security system using the Local Binary Pattern Histogram (LBPH) method for facial recognition, complemented by fingerprint verification as an additional security layer and real-time notifications via the Telegram application. The LBPH method was chosen for its ease of implementation and processing speed, although it has limitations such as sensitivity to lighting changes and potential recognition errors due to similar facial textures. The system utilizes LBPH for initial authentication, followed by fingerprint verification. Users also receive real-time notifications via Telegram to monitor access attempts. Testing showed a facial recognition accuracy of 85% under bright lighting conditions at distances of 30–150 cm, but it decreased to 65% in dim lighting. Fingerprint verification took approximately 2 seconds, while notification delivery required 1–2 seconds on a stable internet network. This system enhances security by ensuring only registered users can unlock the door. If facial recognition fails, the door remains locked without valid fingerprint verification.</strong></p><p class="Standard"><strong><em>Keywords</em></strong> - <em>Face Recognition, Fingerprint Sensor, LBPH, Security System, Telegram Notification</em><em>.</em><em></em></p>