Plant Disease Identification Using Artificial Intelligence and Cloud Collaborative Platform

Abstract

Farmers, consumers, the environment, and the global economy are all at risk from plant diseases. Farmers in India lose 35% of their field harvests to pests and pathogens. Using pesticides without care is also bad for your health because many of them are poisonous and bioaccumulate. You can avoid these bad impacts by finding the disease early, keeping an eye on the crops, and giving the right medicines. Agricultural scientists usually figure out what illness a plant has by looking at its outside signs. But farmers don't have easy access to professionals. Our initiative is the first platform for automated disease detection, tracking, and forecasting that brings all of these things together and lets people work together. With a mobile app, farmers may quickly and correctly find diseases and obtain help by taking pictures of the afflicted areas of plants. The newest AI algorithms for Cloud-based image processing make it possible to diagnose things in real time. The AI model becomes better at what it does by learning from photographs that users upload and suggestions from experts. The platform also lets farmers talk to local specialists. For prevention, a Cloud-based database of geo-tagged photos and micro is used to make illness density maps with predictions on how they will spread. Experts can do disease analytics with geographical visualizations using a web interface. We trained the AI model (CNN) on massive disease datasets made up of plant photos that we took ourselves from several farms over the course of seven months. Plant pathologists checked the results of the automatic CNN model on test photographs. The accuracy of identifying diseases was over 95%. Our solution is a new, flexible, and easy-to-use tool for managing diseases in a wide range of agricultural crops. It can also be used as a Cloud-based service for farmers and professionals to help them grow crops in an environmentally friendly way