基于MATLAB图像处理技术提取火焰高度  

Detection of flame tip height based on MATLAB image processing technology

作　　者：王林[1] 靳红雨[1] 吕东[2] 

机构地区：[1]中国人民武装警察部队学院,河北廊坊065000 [2]公安部天津消防研究所,天津300381

出　　处：《消防科学与技术》2017年第3期366-369,共4页Fire Science and Technology

基　　金：公安部科技强警基础工作专项项目"以水雾为载体的天然气及液化石油气爆炸抑制剂研发"(2015GABJC28)

摘　　要：为避免可视火焰高度因自身周期性脉动和观测者的不同而有所区别,基于MATLAB图像数据处理技术,提出了通过对火焰图像进行灰度化、滤波、画质增强、图像切割和函数化边界线等操作获得火焰实际高度的方法。基于火焰尖端高度提取方法,将图像收集时间延长到一个火焰跳动周期以上,对火焰尖端高度进行算术平均,可得统计意义上的平均火焰高度。开展不同燃气浓度下的点火实验,利用该方法获取本生灯火焰尖端高度,并和光学测高仪读数、Roper模型的理论预测值进行对比,结果显示与实际值匹配良好。To avoid the influence of pulsation frequency and op erators＇ behaviors on visible flame height, based on MATLAB image processing technology, a method to get flame tip height was developed through building fire boundary fitting function. The flame image was under the treatment of graying, wave filtering, image quality strengthening and image segmentation. Based on this method, the collecting time of flame image was prolonged to more than a period of flame pulsation, and arithmetic mean of the heights of flame tips was obtained to get the statistical average flame height. Experiments under different gas concentrations were conducted, the tip height of flame of Bunsen burner were determined, to contrast with the theoreti cally predicted value by Roper model and with the value from optical altimeter. It turns out that the results of this method matches well with actual flame height.

关 键 词：火焰高度 火焰脉动频率 MATLAB图像处理 灰度化 画质增强 光学测高仪 Roper模型 

分 类 号：X924.4[环境科学与工程—安全科学] TP391.4[自动化与计算机技术—计算机应用技术]