基于颜色与运动特征的变电站实时烟雾检测  

Real-time smoke detection in substation based on color and motion features

作　　者：宋伟[1] 王玉平[2] 张文杰[2] 张家齐[2] 郑彦宁[3] 吴高林[1] 

机构地区：[1]国网重庆市电力公司电力科学研究院,重庆401123 [2]重庆大学自动化学院,重庆404100 [3]哈尔滨工业大学深圳研究生院,广东深圳518055

出　　处：《传感器与微系统》2015年第12期127-130,共4页Transducer and Microsystem Technologies

摘　　要：变电站的环境特殊性要求站内火灾检测准确、快速,烟雾检测是早期火灾检测中的常用方法。针对变电站特殊环境,提出一种快速烟雾检测算法。首先,利用改进的Vi Be算法检测变电站实时监控视频中的运动区域,再通过颜色特征确定疑似烟雾区域,然后结合烟雾面积增长特性和移动特性识别烟雾区域。实验结果表明:这种算法能够检测到不同场景中的烟雾,对远距离、稀薄烟雾也能有效检测,满足变电站火灾检测实时性、准确性要求,能够应用于变电站场景烟雾检测。The fire detection in substation is required to be fast and exact,due to its special surroundings,smoke detection is a commonly used method in early fire detection,therefore,a fast smoke detection algorithm is proposed for such unique environment. The specific steps are as follows： Firstly,moving area are determined in real-time by using the improved Vi Be algorithm. Then,the suspected smoke area will be identified by judging color features of the detected regions. Finally,the smoke diffuse and moving characteristics are considered to ascertain the real smoke region. Experimental result shows that this algorithm can be used to detect smoke in different situations,it is not only suitable for nearby smoke detection,but also quite useful for detecting thin smoke in distance. Therefore,this algorithm,which meets the requirements of real-time and accuracy of fire detection in the substation,can be applied in smoke detection in the substation.

关 键 词：变电站烟雾检测 ViBe 颜色特征 面积增长特性 移动特性 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]