A Novel Deep Learning Approach Utilizing HSV Color Space for Real-Time Early Fire

This study proposes a novel fire detection system aimed at improving fire prevention and control efforts, particularly addressing the growing threat posed by fires in urban settings and the alarming frequency of wildfires. Early detection of fire plays a critical role in safeguarding lives and property in places such as shopping malls, offices, factories, and educational institutions. Existing systems for detecting fires, which rely mainly on alarm-based mechanisms, have limitations in their responsiveness. In this project, we developed a machine learning-based fire detection system using computer vision techniques. The proposed model integrates a webcam as part of the surveillance setup to continuously monitor indoor environments. By leveraging video feeds, the system processes the input using machine learning algorithms and the Hue, Saturation, Value (HSV) color space to detect fire at an early stage. OpenCV is employed to handle real-time video processing, ensuring prompt fire detection and timely alerts. This approach demonstrates a significant advancement over traditional systems, offering real-time detection capabilities through the combination of deep learning and advanced color space techniques.