| 摘 要: 针对火灾超早期阶段烟感检测误检率较高的问题，通过融合红外热感视频流、环境烟雾浓度、温度和湿度多种特征，分析易燃物热感视频的特点，利用视频处理方法和图像处理技术实现了检测易燃物中火灾阴燃点的算法。采用“端、边、管、云”物联网体系结构，通过窄带物联网(Narrow Band Internet of Things, NB-IoT)将火警信息发送到物联网公有云。将检测火灾阴燃点的算法部署到嵌入式模组中，实现了在明火出现之前检测到易燃物阴燃点的目的。采集速率可以达到6 帧/秒，检测准确率达到75%。 |
| 关键词: 嵌入式；古建筑；超早期；火灾检测；热成像 |
|
中图分类号: TP399
文献标识码: A
|
| 基金项目: 潮州市科技计划项目(2018GY01)；韩山师范学院博士科研启动项目(QD20190624)；广东省自然科学基金面上项目(2022A1515010990). |
|
| Implementation of Ultra-early Fire Detection System for Ancient Buildings based on Embedded System |
|
MIAO Liming1, QIU Shuwei1, ZHANG Jiexun1, LIANG Fengbin2
|
( 1.Department of Computer and Information Engineering, Hanshan Normal University, Chaozhou 521041, China ;
2.Beijing Dinghan Technology Group Co ., Ltd., Shenzhen 518101, China )
miaolm@hstc.edu.cn; swqiu@hstc.edu.cn; 546572809@qq.com; 811678717@qq.com
|
| Abstract: In view of the high false detection rate of smoke detection in the ultra-early stage of fire, this paper propose to analyze the characteristics of combustible thermal video by fusing multi features of infrared thermal video stream, ambient smoke concentration, temperature and humidity and analyzing. An algorithm of detecting the negative smoldering point of combustible fire is thus realized by using video processing method and image processing technology. The IoT (Internet of Things) system architecture of end-edge-tube-cloud is adopted, and Narrow Band Internet of Things (NB-IoT) is used to send the fire alarm information to the IoT public cloud. The algorithm of detecting the negative smoldering point of fire is deployed in the embedded module to realize the purpose of detecting the negative smoldering point of combustible materials before the appearance of open flame. The acquisition rate reaches 6 frames per second, and the detection accuracy reaches 75%. |
| Keywords: embedding; ancient building; ultra-early; fire detection; thermal imaging |
