摘要
为了提高舰船火灾的识别率,研究了一种基于红外视频图像的火焰特征提取算法。首先对红外视频图像进行预处理,然后利用火焰红色饱和度、亮度值以及红色分量值等判据对疑似火焰区域进行识别,采用四邻域搜索法对疑似火焰区域进行分割,最后根据火焰燃烧初期的特点对火焰特征进行提取。实验结果表明,该算法能够有效分割出疑似火焰区域并提取出火焰特征。
In order to improve the recognition rate of ship fire,a flame feature extraction algorithm based on infrared video image is studied.Firstly,the infrared video image is pre-processed,then the flame red saturation,brightness value and value of the red component criterion was used to identify suspected flame area,the four neighborhood search method for suspected flame region segmentation,and lastly feature extraction of flame according to the early flame characteristics.The experimental results show that the algorithm can effectively segment the suspected flame region and extract the characteristics of the flame.
作者
王凯
冯宪周
WANG Kai;FENG Xianzhou(Naval Representative Office in Zhengzhou Region,Zhengzhou 450015;The 713th Research Institute of China Shipbuilding Industry Corporation,Zhengzhou 450015)
出处
《舰船电子工程》
2018年第3期137-140,共4页
Ship Electronic Engineering
关键词
红外视频图像
火焰识别
特征提取
infrared video image
flame recognition
feature extraction
