The Application Of The Image Processing And Pattern Recognition On Fire Detection Domain

The forest fire is a kind of international and important nature disaster, so many country's governments pay more attention to the forest fire. The forest fire have such features as fire spreading fast, difficulty of extinguishing the fire and rescuing, so it has been an increasing urgent demand that how to effectively protect the safety of forest and detect the fire as early as possible. As a new-style and effective measure for the detection of early fire, image fire detection technology is not affected to the influence of the big space of outdoors and bad environment, so this technology is suited for inspecting the forest fire. In this dissertation, by the aid of digital image processing techniques, several detecting algorithms for forest fire have been designed.The reason of the forest fire happening have been discussed systematically in this dissertation. The difference is also studied between conventional fire detection systems and new-style fire detection techniques. Why the image fire detection technology is suited for inspecting the forest fire and the characteristics of the image fire detection technology have been analyzed in detail as well.The algorithms that the noise filtering, difference image method, the threshold algorithm for image segmentation and their applications in the fire image processing have been studied in this dissertation. Aiming at the problem that the fire video images are difficult to be segmented while the daylight is intensive, a new method is proposed. The integration of these segmenting algorithms can obtain the fire target accurately in the fire image sequence.For the fire image recognition, the features such as color, texture and shape variety of fire flame and other disturbing phenomena have been detailed studied in this dissertation. And the corresponding recognition algorithms have been given. So the early fire flame can be recognized and the disturbing phenomena can be distinguished.The fire detecting scheme based on nerve net has been discussed in the last part of this dissertation. The detail structure of the BP nerve net and the detail design precept of input and output layer have been given. Finally the fire detection system based on BP nerve net has been realized. The experiment results show that the BPNN fire detection algorithm can recognize the forest fire flame effectively and has good anti-interference capability.