林火安防预警与应急处理系统关键技术研究

发布时间：2018-01-31 10:27

本文关键词： 森林火灾 FPGA 红外图像处理　出处：《东华理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：在复杂森林环境中,火灾的发生往往具有很大的突发性和随机性。由于传统的林火安防技术在林火发生初期很难发现火情,因此造成林火蔓延,对周围环境及人畜造成很大的伤害。目前国家在森林防护领域的投入日益加大,对林火的安防技术提出了更高要求。随着可编程嵌入式的发展及红外图像处理技术的成熟,利用红外技术进行林区火情的实时监测成为了研究热点,因此,开发一种可挂载于无人机云台的林火智能监测设备,对监测到的火情进行实时图像处理,这将在早期的火灾控制中起到重要作用。本文通过对森林防火系统的功能和技术架构进行分析,对林火安防领域存在的关键问题进行研究,在FPGA硬件基础上结合红外热像仪,设计了一种体积小、后期可挂载于无人机云台的林火监测系统,其中设计包括了林火红外图像的采集模块、数据缓存模块及显示模块,并针对林火红外图像的特点,使用Matlab对林火红外图像进行了消噪、分割、伪彩增强等图像处理操作,这对后续的火区图像特征提取及识别将起到重要作用。通过在FPGA上运行各种算子卷积模块,林火监测系统在硬件上实现了林火图像的消噪及边缘提取。最后提出了一种林火监测系统与无人机集群灭火技术构成的林火预警与应急处理方案,这将对提高林区防火、救灾的能力起到重要作用。
[Abstract]:In the complex forest environment, the fire often occurs with great sudden and randomness. Because the traditional forest fire security technology is difficult to find the fire in the early stage of forest fire, it causes the forest fire to spread. It has caused great harm to the surrounding environment and people and animals. At present, the national investment in the field of forest protection is increasing day by day. With the development of programmable embedded and infrared image processing technology, real-time monitoring of forest fire using infrared technology has become a research hotspot. A kind of intelligent forest fire monitoring equipment which can be mounted on the cloud head of UAV is developed, and the real-time image processing of the monitored fire is carried out. This will play an important role in early fire control. Through the analysis of the function and technical framework of forest fire prevention system, the key problems in forest fire security field are studied in this paper. Based on the FPGA hardware and infrared thermal imager, a forest fire monitoring system is designed, which is small in volume and can be mounted on the UAV cloud head in the later stage. The design includes the acquisition module of infrared image of forest fire. Data buffer module and display module, and according to the characteristics of infrared image of forest fire, using Matlab to remove noise, segmentation, false color enhancement and other image processing operations. This will play an important role in the subsequent image feature extraction and recognition of the fire zone by running various operator convolution modules on the FPGA. Forest fire monitoring system realizes forest fire image de-noising and edge extraction in hardware. Finally, a forest fire early warning and emergency treatment scheme composed of forest fire monitoring system and unmanned aerial vehicle cluster fire extinguishing technology is put forward. This will improve forest fire prevention, disaster relief capabilities play an important role.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S762;TP391.41

【参考文献】