Rice Disease Prediction Using Machine Learning Techniques

This research Project addresses the critical challenge of rice disease detection and management
faced by farmers in Northern Nigeria through the development of a mobile-based smart agricultural
solution. The system integrates artificial intelligence, image processing, and mobile technology to
provide real-time disease identification and treatment recommendations to farmers. Using a dataset
of over 10,000 rice plant images collected in collaboration with the Kano State Agricultural
Development Agency (KNARDA), the research implemented a Convolutional Neural Network
(CNN) model achieving 94% accuracy in disease detection. The mobile application, developed
using Flutter framework, enables farmers to capture images of affected plants and receive
immediate diagnostic results and treatment recommendations. The system successfully identifies
major rice diseases including bacterial leaf blight, brown spot, and leaf smut, with disease-specific
accuracy ranging from 91% to 95%. Field testing demonstrated a 40% reduction in disease
response time and a 30% decrease in unnecessary pesticide application among test users. This
solution particularly benefits local farmers in Kano State by providing expert-level disease
diagnosis through accessible mobile technology, thereby addressing the limitations of traditional
disease monitoring methods. The research contributes to the advancement of smart agriculture in
Nigeria and provides a framework for implementing similar systems in other agricultural regions.