Application of Image Texture Features in Fire-Detection System

摘要：

A research was performed in this paper that image texture features changed in the course of fire and non-fire. It showed that fire image texture features including energy, entropy, local stability and inertia, would be changed towards an incremental or degressive trend. Before fire happened, the image was steady, and gray distribution was uniformity. The corresponding texture features remained unchanged approximately. When fire happened, gray values of fire image would begin to change because of the movement of smoke and flame. As a result, the gray distribution was no longer even relatively, and texture features would start to change towards a trend. According to analysis of all these information, the mean value of texture wave quantity was calculated to recognize fire sources and sources of disturbances. It has been verified in the experiment that image texture features can be used to discriminate fire from disturbances caused by lamp and candle.

关键词： fire sources image texture features sources of disturbances

作者: CHEN Xiaojuan BU Leping YANG Jiajun

作者单位: College of Electrical and Information Engineering, Naval Univ.of Engineering, Wuhan 430033, Hubei, C 91458Unit of PLA, Sanya 572000, Hainan, China

会议类型: 国际会议

会议名称: The 2010 International Symposium on Safety Science and Technology(2010 安全科学与技术国际会议)

会议地点: 杭州

会议语种:英文

页码: 841-846

在线出版日期: 2010-10-26（万方平台首次上网日期，不代表论文的发表时间）