微型无人机林火监测系统的设计与实现  

Design and implementation of a forest fire monitoring system using a miniature unmanned aerial vehicle

作　　者：马瑞升[1,2] 杨斌[3] 张利辉[4] 刘志平[1,2] 

机构地区：[1]广西气象减灾研究所,广西南宁530022 [2]国家卫星气象中心遥感应用试验基地,广西南宁530022 [3]淮阴师范学院,江苏淮安223300 [4]中国空气动力研究与发展中心,四川绵阳621000

出　　处：《浙江农林大学学报》2012年第5期783-789,共7页Journal of Zhejiang A&F University

基　　金：中国气象局气象新技术推广项目(CMATG2007M40);中国气象局气象新技术推广项目(CMATG2004M22);国家高技术研究发展计划("863"计划)项目(2006AA12Z125)

摘　　要：为探索林火监测的新手段,提高森林安全管理水平,将实时视频传输和计算机技术相结合,研制了林火监测任务设备,并与无人机飞行平台共同构成了微型无人机林火监测系统。以林区火场影像为样本,像素RGB值为分类依据,使用统计产品与服务解决方案软件进行聚类分析,得到基于颜色特征的烟雾识别模型。该模型经过光谱特性改进和消噪处理后,对于试验影像的烟雾识别率在77%以上。利用该模型,系统可在飞行中实时完成对地面火情的探测并给出报警提示。在飞行试验的基础上,提出无人机林火监测系统的飞行作业流程,并讨论了系统运行的经济性和目前存在的问题。无人机林火监测系统同样可用于林区气象探测、可燃物和病虫害调查等方面,切实提高林区管理的技术水平。To explore a new means of forest fire monitoring and to improve the management tevet oI ~orest safe ty, realtime video and computer technology were combined to develop a new piece of forest fire monitoring e quipment. This equipment and a miniature unmanned aerial vehicle （UAV） together formed the miniature aerial forest fire monitoring system. Sampled RGB value of image pixels, a smoke recognition model was established by using the statistical product and service solution （SPSS） for cluster analysis. The model had been improved for smoke spectrum and noise reduction, cloud recognize more than 77% smoke pixels of test images. With the recognition model, immediate fire detection and warning could occur. Based on flight tests of the system, this study proposed operational processes, analyzed running costs and existing problems at present. In addition, ap plication of a UAV forest fire monitoring system as well as meteorological sounding, combustible matter moni toring, pest and disease survey and others, could enhance forestry management. [Ch, 5 fig. 2 tab. 14 ref.]

关 键 词：森林保护 无人机 森林火灾 监测 识别 模型 

分 类 号：S762.2[农业科学—森林保护学]