A FAST AND EFFICIENT VISION BASED APPROACH FOR DETECTION OF FIRE IN REAL TIME SCENARIO

Fire is a huge serious disruption which leads to economic and environmental losses. So it is necessary to detect occurrence of fire at early stage. Alarm is not issued unless particles reach the sensors to activate them and requires large response time. Also, infrared and ultraviolet sensors which are commonly used produce many false alarms. With the help of computer vision techniques, it is possible to get better results than conventional systems because images can provide more reliable information. In this paper OpenCV library is used to process video images more accurately with fast speed. Also an innovative approach is proposed, which is based on video processing to overcome drawbacks of sensor based fire detection methods. The main consideration in this proposed technique is that, it is based on three algorithms mainly used for flame detection & they are motion detection algorithm, edge detection algorithm and colour detection algorithm. It is observed that this approach based on three techniques gives better results then existing fire detection techniques.