Face Recognition and Tracking System for Rescue Operation in Fire Emergency

Abstract

Fire incidents remain one of the major threats to human life and property, making effective rescue operations critical during emergencies. This project presents the development of a system designed to assist rescue workers in identifying and tracking individuals inside a building during a fire. The system integrates an ESP32-CAM module for real-time video streaming and facial recognition, combined with motion (PIR) and smoke (MQ2) sensors to detect human presence and potential hazards. Recognized faces and sensor data are transmitted to Firebase Realtime Database and displayed on a mobile application to provide live updates for rescue teams. This real-time monitoring enables responders to locate trapped individuals more quickly, even in smoke-filled environments where visibility is limited. The system complements traditional rescue methods by providing additional information to improve response speed and accuracy. Limitations include reduced face recognition accuracy in crowded scenes and lower performance under poor lighting conditions. Overall, the system demonstrates promising potential to enhance the effectiveness of fire rescue operations.