摘要
根据船舶火灾探测的特点以及火灾早期预报的要求,设计了带图像信息的火灾自动报警系统,并将自适应加权融合算法应用于该系统。在该系统中信息层完成数据采集、处理,特征层运用自适应加权融合算法进行特征融合,决策层运用概率神经网络进行决策融合,完成船舶火灾探测系统的自动报警输出。实验表明,将该算法应用于此系统能够准确、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。
According to the demanding of fire detection on shipboard,a fire detection system with image information was designed.And the self-adaptive weighted fusion algorithm was used in this system.In the information layer of fire detection system,the data about texture feature of fire image,temperature and smoke thickness was collected and pretreated.In the feature layer,self-adaptive weighted fusion algorithm was applied in the feature layer.And in the decision layer,the PNN（probabilistic neural networks（PNN）） algorithm was used to make the decision fusion in order to get the output of the fire detection system in the ship.It has been verified in the experiment that this system can distinguish fire from nuisance sources caused by daylight lamp and alcohol burner.And it has a better anti-jamming performance.
出处
《舰船电子工程》
2011年第3期52-55,共4页
Ship Electronic Engineering
基金
国家自然科学基金(编号:50677069)资助
关键词
自适应加权
数据采集
数据融合
干扰源
self-adaptive weighted fusion
data collection
data fusion
anti-jamming performance
