摘要
针对火灾探测系统中普遍存在的漏报警和误报警等抗干扰性差的问题,设计了带图像信息的火灾自动报警系统。该系统中信息层完成火灾现场图像纹理、烟雾浓度、温度的数据采集、处理,特征层运用多种数据融合算法分别进行火灾识别,决策层分析3种识别结果的输出组合方式,完成重要场所火灾系统的自动报警输出。实验表明,该系统能够准确、快速探测火灾,识别日光灯、酒精灯等干扰源,具有较强的抗干扰性,同时具有良好的经济效益。
In order to reduce false alarm and miss alarm failures of fire detection system,a fire detection system with image information was designed.In the information layer of the fire detection system,the data about the texture feature of fire image,temperature and smoke thickness are collected and pretreated.In the feature layer,several data fusion algorithms are used to recognize the fires.And in the decision layer,three kinds of recognition results are analyzed to get a combined output and a fire alarm output for important locations is given.Experimental results indicate that this system can detect fires accurately and quickly and can distinguish fires from nuisance sources caused by fluorescent lamp and alcohol burner and etc.Therefore,the proposed system has good anti-jamming performance and good economic benefits.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2010年第12期2848-2853,共6页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(50677069)资助项目
关键词
火灾报警系统
数据采集
数据融合
抗干扰性
fire detection system
data collection
data fusion
anti-jamming performance
