摘要
对无人机设计方案、图像处理和火焰分割算法的技术原理进行了介绍,并利用BP神经网络对图像中的火焰面积变化率和火焰尖角等特征进行识别,实现了对森林火灾的快速监测。实验结果表明:系统的准确率为98.5%,比普通神经网络的84.5%更高;耗时仅22 s,比普通神经网络159 s缩短很多。这表明,BP神经网络是更可靠且更有效率的火灾识别方案。
First introduces the UAV design scheme,image processing and the technical principle of flame segmentation algorithm,and then uses BP neural network to identify the characteristics of the flame area change rate and flame sharp angle in the image,so as to realize the rapid monitoring of forest fire.The experimental results show that the accuracy of the algorithm studied in this paper is 98.5%,which is higher than 84.5%of the ordinary neural network,and takes only 22 seconds,which is much shorter than 159 seconds of the ordinary neural network,which proves that the BP neural network is a more reliable and efficient fire identification scheme.
作者
杨静
Yang Jing(Hubei University of Police,Wuhan 430034,China)
出处
《农机化研究》
北大核心
2025年第2期205-209,共5页
Journal of Agricultural Mechanization Research
基金
湖北省电子取证及可信应用协创中心项目(HBDZQZ001)
湖北警官学院2023年度院级重点科研项目(HJ2023ZD01)。
关键词
森林防火
无人机
图像处理
BP神经网络
forest fire prevention
UAV
image processing
BP neural network
作者简介
杨静(1979-),女,湖北孝感人,副教授,硕士,E-mail:15190137@qq.com。
