IMAGE-TYPE Fire Flame Detecting And Alarming System

The aim of this thesis was to synthesis the processes of the project, which is making a detecting and alarming system by a popular color vidicon and DSP technology.The paper presents a real-time system for automatic fire detecting and alarming using color video input. Such a system has many important commercial applications and has significant advantages over traditional smoking, temperature, ultraviolet and infrared image fire detectors. These advantages include improved detection, fewer false alarms, and additional descriptive or display information about fire location, size, and growth rate. From the physical optical properties of fire, we derive algorithms for fire detection based on the energy spectral, color distributing, spatial, and temporal properties of fire events. These algorithms have been integrated into a system that has been model successfully on a wide range of fire, an over speed respond of fire and the lowest false alarm stimuli. At last I present experimental project demonstrating system performance on a burning jet fuel fire, a test fire of EN54-9 and Q/3200GKV20-2003.