摘要
基于常用的火灾探测方法中都是使用单一特征,对火灾的图像进行判别,这一技术在复杂的场所中并不能有效地去除干扰源,从而出现误报、漏报等问题,提高火灾探测的有效性和准确性的目的。在对火焰的颜色模型、颜色特征和形状特征进行研究的基础上,采用数字图像处理等方法,通过提取火焰的一阶颜色矩和圆形度等多种特征,并且经过支持向量机分类器进行分类,从而对火灾火焰和干扰源进行判别的试验,得出这一方法可以很好的排除干扰源信息,探测率高,可靠性好的结论。
Since most video based fire detection techniques use a single feature for fire detection, interference in complex fire conditions cannot be effectively removed, which may severely affect the accuracy of detection. A new video flame detection algorithm using multi-feature fusion method is developed in this paper. The color and morphological features of fire flames,including the first order color moments of the flame and circular degree, are analyzed in detail. A support vector machine（SVM） based classifier is then used to process the features and recognize fire flames from video images that may contain interference sources. From the testing results, it can be seen that the method can effectively remove interference information,and thus have high detection rate and good reliability.
出处
《电子设计工程》
2016年第21期188-190,共3页
Electronic Design Engineering
基金
江苏科技大学博士科研启动基金资助项目(635301202)
关键词
颜色模型
数字图像处理
火焰特征
支持向量机
color model
digital image processing
flame features
Support Vector Machine（SVM）
