Study And Implementation Of Rice Disease Diagnoses Expert System

As one of the most important grains in our country, rice suffers a lot from various diseases which will cause the drop in production and harm the safety of our yield directly. Yet, in our country, there are a large number of people working on rice-plantation who are poorly-educated; correspondingly, the number of people engaged in actual plant protection work is so limited that it is difficult for rice-planters to diagnose, prevent and cure different kinds of diseases. Therefore, the development of an expert system for rice disease diagnoses has its significance in guiding rice-planters to apply pesticides and fertilizers, prevent and cure rice diseases and improve its production.This study combines AI with computer vision, develops an expert system for rice disease diagnoses which has the functions of intelligent reasoning diagnoses and image recognition diagnoses by using Microsoft SQL Server and Borland C++ Builder, elaborates the WWW and WAP version by using ASP and WML language which makes the users access this system in more flexible way and simplifies the way of spreading the system so that to improve the scope of application.This study tries a new way developing an expert system for rice disease diagnoses which combines AI with computer vision, brings up Knowledge representation method that using production rule based on mark, concludes a characteristic parameters system marking the information of disease spot and contrasts the effects between Fuzzy weighted and BP Neural Network in classification of rice diseases images.