Variance Feature based Fuzzy Inference System for Diagnosis of Rice Plant Diseases

Timely diagnosis of diseases in the crop is very important for Integrated Pest Management (IPM). Still, in many parts of the world, diseases identification is based on human expertise by naked eye observation. This observation and perception varied from person to person impact survivability of the plant and control of diseases. Therefore, farmer and government agencies required automated Plant Diseases Decision Support System (PDDSS) to enforce IPM during farming for diagnosis of the diseases. In this research, an expert system based on Sugeno fuzzy inference system had been developed to identify diseases based on symptoms appear in various parts of the plant. In this research, Rice plant had been considered with four categories of studies; three diseases infected rice crop and one non infected crop. Each disease infected or non-infected rice plant reflects unique symptoms appear in the plant leaf, root and node. These unique features had been extracted with the help of image processing and variance analysis had been done. The variance features had been used to design Sugeno fuzzy model as an expert system to the diagnosis of the diseases. The training and testing results show the efficient and accurate result.