AgriAid: Android-based Crop Disease Detection and Advisory System

Agriculture faces mounting threats from plant diseases that significantly reduce yields and incomes for smallholder farmers. AgriAid is an Android-based application that provides offline, on-device crop leaf disease diagnosis using multiple optimized TensorFlow Lite (TFLite) models. The application supports several crops and executes real-time inference on resource-constrained devices, delivering high accuracy and low latency under varied field conditions. This paper presents the system design, model integration approach, implementation details, and empirical performance metrics obtained through extensive testing.