基于DCNN的火灾探测与定位系统  

Fire Detection and Location System Based on DCNN

作　　者：张怡[1] ZHANG Yi(The Engineering&Technical College,Chengdu University of Technology,Leshan 614000,China)

机构地区：[1]成都理工大学工程技术学院,四川乐山614000

出　　处：《微型电脑应用》2021年第5期58-61,共4页Microcomputer Applications

基　　金：四川省教育厅科技项目(17ZB0062);乐山市科技局(19SZD106)。

摘　　要：由于卷积神经网络在机器视觉中表现出色,因而其应用范围越来越广。为了提高火灾探测系统中的探测精度,减小火灾的负面影响,越来越多的基于机器视觉的火灾检测方法被提出。但由于现有的方法存在较高误报率,准确度较低等缺点。为了解决这些问题,提出了一个基于DCNN的火灾检测系统。这个系统使用较小的卷积核,并且不包含密集的、完全连接层,保证了计算量较小。实验结果表明,提出的解决方案可以达到与其他更复杂模型相当的精度。并且在火灾定位中表现较好。Because convolution neural networks perform well in machine vision,its application range is getting wider and wider.In order to improve the detection accuracy of the fire detection system and reduce the negative impact of fire,more and more fire detection methods based on machine vision have been proposed.However,the existing methods have shortcomings such as high false alarm rate and low accuracy.In order to solve these problems,a DCNN-based fire detection system is proposed by inspiration of the SqueezeNet architecture.The system uses a smaller convolution kernel,and does not contain dense,fully connected layers,which ensures a small computational requirement.Experimental results show that the proposed solution can achieve accuracy comparable to other more complex models.And it performs better in fire positioning.

关 键 词：卷积神经网络 火灾定位 火灾检测 深度学习 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]