KisanCare: An AI-powered System for Crop Disease and Nutrient Deficiency Detection Using Convolutional Neural Networks

Abstract

This study presents KisanCare, an AI-powered web application designed for early detection of crop diseases and nutrient deficiencies from leaf images and soil parameters. The system employs a custom-trained Convolutional Neural Network (CNN) on a dataset of over 55,000 images spanning 32 classes, enabling accurate recognition of plant health conditions. In addition to image-based classification, an advanced mode integrates rule-based nutrient interaction analysis to provide preventive recommendations. The multilingual user interface, developed in Flask with English, Hindi, and Telugu support, ensures accessibility for farmers in diverse regions. Evaluation results demonstrate high prediction accuracy, supported by confidence scores and actionable treatment advice. The proposed solution bridges the gap between academic AI research and practical precision agriculture, empowering farmers with a cost-effective, scalable, and user-friendly diagnostic tool.