Deep Learning-Based Fire and Smoke Detection System
B. Senthilnayaki, M. Anousouya Devi, S. Abijah Roseline, P. Dharanyadevi
Abstract
This paper presents a novel Deep Learning-Based Fire and Smoke Detection System designed to surpass conventional sensor-dependent methods. Unlike traditional fire detection systems reliant on sensors, this system utilizes computer vision techniques and deep learning algorithms, specifically convolutional neural networks, to analyze video footage and images for the presence of fire or smoke. The proposed system addresses the limitations of sensor-based approaches, making it particularly suitable for extensive environments like large forested areas or open spaces where sensor deployment is impractical. By leveraging the power of deep learning, the system enhances the accuracy and efficiency of fire and smoke detection, enabling it to adapt to complex real-world scenarios. The core objective is to reduce the frequency and severity of fire accidents. Upon detecting fire or smoke in a monitored area, the system triggers immediate notifications via email to the relevant user and activates a fire alarm. This comprehensive approach enhances the capabilities of video surveillance systems in forest settings and contributes to a more proactive and responsive fire prevention strategy.