Abstract
In agriculture-dependent countries like Malawi, crop diseases significantly threaten food security and farmer livelihoods, particularly among smallholder farmers who often lack timely access to expert agronomists for accurate diagnosis. This project addresses that gap by developing a mobile application that leverages machine learning and image processing to detect crop diseases from images of plant leaves. Using convolutional neural networks (CNNs), the app analyzes visual features such as color, texture, and shape to identify common diseases affecting key crops. Designed with a simple, mobile-friendly interface, the app enables farmers to capture and upload images directly from a smartphone, receiving instant, reliable diagnoses along with basic guidance for treatment or disease management. By facilitating rapid detection and informed decision-making, the app empowers farmers to act promptly, minimize crop losses, and adopt more sustainable farming practices. This solution not only strengthens agricultural productivity and resilience but also contributes to improved food security and economic stability in rural communities across Malawi and similar regions. Moreover, the app is designed to function even in low-connectivity environments, ensuring accessibility in remote areas. It incorporates local language support and visual cues to accommodate farmers with limited literacy, making it a truly inclusive tool. Over time, the system can be expanded to include a wider range of crops and diseases, and its underlying database can be enhanced through user-contributed images, continuously improving the model’s accuracy. By integrating technology with traditional farming, this innovation has the potential to transform agricultural practices and uplift the livelihoods of thousands of farmers.