Deep Learning-Based Web Application for Crop Disease Detection Using CNN and Streamlet

Abstract

Crop diseases can cause significant yield losses and threaten food security if not diagnosed and treated in time. In this paper, we present a real-time web application for crop disease detection using deep learning techniques, primarily Convolutional Neural Networks (CNNs). The application allows farmers to upload leaf images through a Streamlet interface, which are then analyzed by a trained CNN model to detect diseases. The system provides fast, accurate predictions along with suggestions for treatments, promoting smart and sustainable farming. For scalability and maintainability, it makes use of a modular backend, picture preprocessing methods, and the Plant Village dataset. Performance is evaluated using accuracy, precision, recall, and user feedback. Key Words: Crop Disease Detection, Convolutional Neural Network, Deep Learning, Streamlit, Smart Farming, Plant Village, AI in Agriculture

Code and data availability

The paper uses the public PlantVillage dataset, but that is a cited third-party resource, not a paper-specific asset. No author code, trained model, or data deposit with a public URL is mentioned; the only URL (Streamlit docs) is a generic library reference.

No evidence-backed public reproduction asset is currently recorded.