Plant Disease Classification System ( ML-Powered Agricultural Solution )
Benjamin Koomson
Published
Developed a machine learning–powered agricultural web application that classifies plant diseases from leaf images using a Convolutional Neural Network (CNN) built with PyTorch. Integrated the model into a Flask-based backend API to handle image uploads and return real-time classification results. Implemented a responsive web interface for users to view disease predictions, check weather conditions via an external API, and receive text-to-speech output using ElevenLabs for accessibility. Utilized SQLAlchemy ORM to manage structured data storage, including user inputs, predictions, and weather data. Led a 3-person development team, coordinated sprints, and contributed to project documentation covering API design and deployment procedures.