一种基于飘动性分析的视频烟雾检测新方法  

Flutter Analysis Based Video Smoke Detection

作　　者：王涛[1] 刘渊[2] 谢振平[2] 

机构地区：[1]江南大学物联网工程学院,无锡214122 [2]江南大学数字媒体学院,无锡214122

出　　处：《电子与信息学报》2011年第5期1024-1029,共6页Journal of Electronics & Information Technology

基　　金：江苏省科技支撑计划项目(BE2008009)资助课题

摘　　要：该文通过对烟雾飘动特性的分析,提出一种基于视频图像的火灾烟雾检测新方法。首先利用滑动时间窗分析视频中运动区域的飘动性特征,提取出飘动方向、周期飘动强度、周期有效飘动强度和周期逆向飘动强度等特征向量;然后使用神经模糊推理系统训练并检测烟雾。与利用其它特征检测烟雾算法的对比实验表明,该算法具有较好的ROC(Receiver Operating Characteristic)曲线特性,具备较高的实际使用价值。Based on the flutter characteristic of smoke,a novel video smoke detection algorithm is proposed.Firstly,the motion region is detected from video frames.Then four flutter features of the motion region are extracted over a sliding time window,including the flutter direction and three types of flutter intensities.Finally,the neuro-fuzzy inference system is used to detect smoke,in which fuzzy rules and membership functions are trained according to the valid sample set.Experimental results show that the new algorithm has excellent performance on Receiver Operating Characteristic（ROC） curve which is very important for real application.

关 键 词：图像处理 视频烟雾检测 飘动性分析 神经模糊推理系统 ROC曲线 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]