小麦病虫害便捷识别模型和远程诊断系统  

Simplified identification model and an expert system for long-distance diagnosis of wheat diseases and pests

作　　者：马春森[1] 何鹏[2] 马罡[2] 杨和平[2] 

机构地区：[1]中国农业科学院植物保护研究所,北京100081 [2]中国农业科学院农业环境与可持续发展研究所,北京100081

出　　处：《植物保护》2011年第3期165-169,共5页Plant Protection

基　　金：国家高技术研究发展计划("863"计划)(2002AA241231)

摘　　要：以小麦病虫害为对象,使用通俗易懂的语言和大量图片,构建了小麦病虫害文字数据库、图库,并结合便捷识别模型和网络技术集成开发了小麦病虫害远程诊断系统。全面收集整理我国小麦病虫害种类名单,病害35种,虫害42种;根据小麦所处的生育阶段和为害部位排除名单中大多数不可能的种类;以病虫害生命活动在作物和环境中留下的直观痕迹和粗略的外形特征,鉴别剩下的可能种类;必要时以虫体或病原菌本身的细微形态特征为依据,确诊病虫害种类。系统应用测试结果表明,准确率达90.6%,适合专业人员和普通农民使用。To meet the urgent demands of farmers and agricultural extension service, we developed a simplified identification model to identify wheat diseases and pests. Firstly, all available information on wheat pests and diseases in China was collected. A name list totally including 35 diseases and 42 pests was work out. Secondly, most diseases and pests in the name list were excluded by comparing their occurrence phenology and damaged plant organs. Thirdly, the rest was differentiated by comparing symptoms, traces remained by insects and significant morphological characteristics. At last, if necessary, the exact species of insects or phytopathogens are identified by observing the detailed morphological characteristics. We constructed a database that included detailed information on morphology, biology, ecology and control technology of the disease and pests described in texts and pictures. We developed a longdistance diagnosis system by incorporating the simplified identification model and database with internet technology. The system was tested and its accuracy reached 90.6%. The system is easy to be used by experts of plant protection, most farmers and extension workers.

关 键 词：识别模型 网络专家系统 小麦 病虫害 

分 类 号：S43[农业科学—农业昆虫与害虫防治]