Research On Flame Detection Based On Video Image Processing

With more fire accidents breaking out, it has brought immensely material and spiritual loss to the people so that it’s difficult to detect fire accident of large area real-time by using conventional fire detection system. Traditional sensor-based fire detection system will tend to fail to detect fire accidents by analyzing the physical phenomenon and/or the features of chemical reaction products because it is easily affected by the equipments themselves, places and surroundings. As a consequence, how to detect fire accidents fast and accurately have become many people’s common concerns.In order to solve the problem of real-time performance of fire detection system, a fire flame detection method based on video image processing was put forward to substitute the conventional fire detection techniques. The proposed fire detection system has an extremely extensive application prospect in the market. First of all, the method of Ohta color space combined with background updating was utilized to extract the moving object based on flame color and movement characteristics. It could segment the moving targets of flame color features and rule out a large number of non- flame moving targets. Meanwhile, it could reduce a lot of noise compared with the moving objects extracted from other background updating, helping facilitating subsequent rapid processing. Secondly, according to the characteristics of flame propagation, this paper used the method of incremental binary for moving targets, and then averaged, after that segmented the lighter pixel area through binarization method so as to rule out quick moving goals and retain the flame area of slow movement. Next, we could get moving area conforming to flame color based on the flame color and its movement characteristics. Finally, the corner feature is used to describe the characteristics of flame random movement. Testing the flame characteristics through the above methods, we could eventually determine the flame area so as to give a warning.The experimental results show that the flame detection system based on video image processing for fire monitoring research can have very good real-time monitoring, for further research on a wide range of fire, all-round monitoring is of great significance.