Technique for automatic forest fire surveillance using visible light image

摘要: Fires are one of the major hazards to forest. Governments various countries spend an enormous amount manpower and financial resources prevent forest fires every year. Many research projects, which monitor identify fires, have been developed by using infrared remote sensing. In this paper image features all kinds fire background surface objects on visible light analyzed. On base spectrum feature in spectrum, feasibility method recognize is discussed, then a new with put forward fires. Firstly, serial frame generalized, establishes theoretical foundation for further recognition. Secondly, suspicious area found single according difference got from first step. Thirdly, some erroneous judgments eliminated response continuous frames. end Jinggang Mountain locates Jiangxi province China taken as example. More than 30 segments flow, photographing different climate angles used fire. Conclusion can be drawn that rate recognition 100%, accuracy 85%.