Study On Fire Detection Method Based On Particle Image Analysis

Xueming Shu,Hongyong Yuan,Jun Fang,Guofeng Su
2004-01-01
Abstract:Smoke particle is one of the important parameters of fire detection, and particle image is the important characteristic of smoke particles. In this paper, particle images of fire smoke were acquired by Field emission Scanning Electron Microscopy (SEM), and processed by a certain image processing method. From this work, the particle's fractal dimension and prefactor of the fire smoke of cotton wick, polyurethane and beech wood were obtained. Then, in order to get the particle size and its distribution of fire smoke, a new method of laser sheet imaging-based was developed. A set of experimental apparatus to capture smoke particles were designed, which included three parts: the smoke passage system, light scattering system and particle imaging system. From the experimental apparatus and image processing algorithm, particle images, particle sizes as well as their distributions of different fuels were acquired. So, this method provided us with not only the smoke particle morphology but also its size and distribution, which maybe benefit for the research of smoke detection technology.
What problem does this paper attempt to address?