An image processing technique for fire detection in video images

被引：194

作者：

Marbach, G ^{[1
]}

Loepfe, M ^{[1
]}

Brupbacher, T ^{[1
]}

机构：

[1] Siemens Switzerland Ltd, CH-8708 Mannedorf, Switzerland

来源：

FIRE SAFETY JOURNAL | 2006年 / 41卷 / 04期

关键词：

video fire detection; real time image processing; pattern recognition;

D O I：

10.1016/j.firesaf.2006.02.001

中图分类号：

TU [建筑科学];

学科分类号：

0813 ;

摘要：

This paper presents an image processing technique for automatic real time fire detection in video images. The underlying algorithm is based on the temporal variation of fire intensity captured by a visual image sensor. The full image sequences are analyzed to select a candidate flame region. Characteristic fire features are extracted from the candidate region and combined to determine the presence of fire or non-fire patterns. Fire alarm is triggered if the fire pattern persists over a period of time. (C) 2006 Elsevier Ltd. All rights reserved.

引用

页码：285 / 289

页数：5