A mobile-based crop disease detection system that uses a lightweight deep learning model (MobileNetV2) to identify plant diseases from leaf images and provide instant, offline advisory (symptoms, treatment, and prevention) to farmers, with optional cloud-based support for advanced guidance when internet is available.
KrishiCare: AI-Based Crop Disease Detection and Advisory System
KrishiCare is a mobile-based system that helps farmers detect crop diseases using leaf images and receive immediate guidance. It uses a lightweight deep learning model (MobileNetV2) optimized with TensorFlow Lite, allowing fast and efficient on-device prediction without requiring internet access.
The user captures or selects a leaf image, and the model identifies the disease along with a confidence score. This prediction is then linked to a structured offline dataset that provides key information such as symptoms, treatment methods, and preventive measures in a simple, actionable format.
The system works in two modes:
Offline Mode: Disease detection and advisory using local model and dataset
Online Mode (optional): Advanced support using cloud-based AI when internet is available
KrishiCare is designed for low-end smartphones, making it suitable for rural areas with limited connectivity. It enables early disease detection, reduces dependency on experts, and helps farmers take timely action to protect crop yield.