基于图像规则推理的玉米病虫草害诊断系统的设计  

Design of an expert diagnosis system for diseases,pests and weeds of corn base on image rule-based reasoning

作　　者：刘同海[1,2] 黄斌博[1] 李少昆[3] 滕光辉 李乃祥[1] 何玉香[1] 

机构地区：[1]天津农学院计算机科学与信息工程系,天津300384 [2]中国农业大学水利与土木工程学院,北京100083 [3]中国农业科学院作物科学研究所,北京100081

出　　处：《中国农业大学学报》2012年第4期154-158,共5页Journal of China Agricultural University

基　　金：国家星火计划(2010GA610018);天津市自然科学基金(08JCYBJC02200);天津农业科技推广与转化项目(0804110)

摘　　要：为解决因植保人员短缺而使玉米病虫草害不能及时得到诊断与防治,导致玉米生产减产乃至绝产的问题,本研究基于不同发病部位的病症及虫害、草害的图像知识和文本知识,采用二叉树检索算法,设计开发了玉米病虫害诊断专家系统。实际应用结果表明:系统可实现对病虫草害图像、文本和图像推理规则等知识的管理及病虫草害智能推理诊断等功能;可实现对虫害草害实现指认式诊断、解释推理过程和推理结论。通过该系统推出的玉米病虫害诊断结论,农户可针对病虫草害种类及危害程度及时防治病虫害,减少玉米生产损失。In order to solve the production reduction of corn,or even no harvest,caused by lack of timely diagnosis and cure of diseases, pests and weeds, for the shortage of plant protection professionals, an expert diagnosis system was developed by using the search algorithm on binary tree based on the symptom of the diseased part of the plant and the image and text knowledge of pests and weeds. The application result shows that the system can realize management of the image knowledge of pests and weeds, the management of the text and image knowledge of rule-based reasoning, and the intelligence diagnosis of plant diseases, pests and weeds. The system also can accomplish the identifying recognition and interpret the process and results of reasoning. By using the diagnostic conclusions of diseases, pests and weeds,farmers can prevent and cure the diseases in time to reduce the foss of corn yield.

关 键 词：玉米病虫草害 专家系统 图像推理诊断 二叉树 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]