Nowadays, the assurance of fire prevention in buildings is a matter of great concern. The reason comes from practice, there have been incidents originating from fire prevention that have led to serious consequences in society. Up to now, researchers have focused on detecting fires in forests, in large areas, but very few studies have focused on detecting and predicting fires in buildings. For that reason, this study focuses on camera-based fire detection and localization problem in buildings which is applicable to the alerting system implemented inside. This approach utilizes the You Only Look Once (YOLO) v3 algorithm combined with a Convolutional Neural Network (CNN) to early detect and assess fire rick for indoors alert. The results showed that the proposed solution is promising for decision making and early handling fire beforehand to ensure the safety inside buildings.
Năm 2024 
