Study On The Technology Of Large Space Fire Detection Based On Image Processing

For the fire's characteristics of spreading quickly in large space and difficulty of fighting and rescue, it is very important to detect and guard effectively in early stage. Generally, the fire detecting system is based on infrared detection technology in recent years.In this thesis a fire detecting system based on image processing has been developed: the image capture experiment system has been built, and the massive images of fire and incandescent lamp are gathered for test analysis. By analyzing different images, method of difference image is used to obtain suspicious flame region image. Average value filter method is used to remove the noise of image and fixed threshold is used to implement binarization images. The fire flame is distinguished by the morphological characters of the area change and similarity of suspicious flame region which are from adjacent images, as well as the circularity of suspicious flame region in each image. The methods for getting centre coordinate of flame in fire image and for the coordinate transformation between fire image and construction are given.This method has already been applied preliminarily in a project, and is validated to be basically feasible.