Study On Soybean Diseases Expert System Of Heilongjiang Reclamation Area

In order to solve issues that the soybean disease diagnosis is not high degree of automation and not timely in Heilongjiang reclamation area, this study combines plant protection knowledge with modern information technology, to put forward a method of intelligent diagnosis of soybean diseases, to develop expert system for soybean disease Heilongjiang reclamation area, to realize soybean leaf disease diagnosis on the remote, large area, fast, intelligent. Soybean growers can diagnose and control common disease of soybean leaf effectively, thus reducing disease losses, improving soybean yields, reducing pesticide inputs , providing technical support for soybean production and management.In this paper, author analyzes actuality, shortage and tendency of technology about crop diseases intelligence diagnosis and study about expert system overseas, proposes an idea about combination of crop disease intelligent diagnostic techniques and soybean expert system. Author introduces the main source of the soybean leaf disease, collection tools and collection methods, analyses soybean leaf disease expert system technical feasibility and economic feasibility, confirms system basic function by the method of requirement analysis, develops system operating platform by C/S model. Author puts forword technology measures and design programs for soybean disease expert system by the theory and methods of knowledge base and database.Appling digital image processing, artificial neural networks and intelligent pattern recognition technology to the key technology of soybean disease diagnosis. Establishing three-layered neural network model for soybean disease segmentation, extracting and measuring shape features, color feature and texture feature of the lesion area, establishing a disease recognition model. Presenting the best histogram entropy method by improving genetic algorithm in order to solve the problem of weeds segmentation, analyzing the knowledge base model of soybean insect pests, and bringing forword the method of inset pest diagnosis based on reasoning mechanism. Designing soybean disease expert system function and database, constructing knowledge system and expert consulting system with space-time adaptive. Using Delphi programming language, using C / S development model, using SQL Server 2000 database management tool, to develop soybean disease expert system. System major functions realize soybean varieties resource management, insect resource management, disease resource management, weeds resource management, soybean common pesticides resource management, cultivation techniques, program design expertise and system maintenance management.