Smart Detection of Rice Leaf Disease using Deep Learning Models

The worldwide consumption of food depends heavily on rice which serves as one of the essential crops required to sustain people worldwide. The agricultural production of rice faces difficulties from leaf diseases such as brown spot and leaf smut which significantly affect food security through yield reduction. Leaf disease detection through conventional methods requires professional expertise and manual labor because services are limited in rural areas. This problem required a deep learning solution using the MobileNetV2 model to achieve precise diagnosis of rice leaf diseases effectively. The system underwent training with multiple image classes featuring healthy rice leaves alongside brown spot and leaf smut diseases. The MobileNetV2 model delivered 98% classification accuracy through its light processing speed on mobile and edge devices. The system functions in realtime so it meets the requirements to run via mobile applications in agricultural fields. Using this methodology farmers gain access to an economical detection system which tracks crop health and helps them take prompt actions leading to better rice cultivation sustainability and enhanced yields.