Cacao plant leaf disease detection and classification
Emmanuel Tovurawa , Blessing Dlamini
Partner: cape-coast
Year: 2024
Abstract:
Cacao farming is a crucial revenue-generating agricultural export in Ghana, but the prevalence of disease-infected cacao plants poses huge risks to crop yields and farmers' incomes. Leveraging deep learning techniques offers a promising solution for early detection of cacao plant diseases. This project encompasses a comprehensive approach, starting with Exploratory Data Analysis (EDA) of a dataset consisting of healthy and diseased cacao plant images from Ghanaian farms. EDA helps identify patterns and understand dataset characteristics, laying the foundation for robust machine learning models tailored to the specific challenges of cacao farming in Ghana. We develop and evaluate deep learning models for detecting and categorizing cacao plant diseases, utilizing data from the University of Ghana and other open-source datasets. These models are designed with the Predictability, Compatibility, and Stability (PCS) framework in mind, ensuring reliability and effectiveness in disease detection. Moreover, we emphasize the importance of model deployment, enabling farmers to upload images and receive real-time diagnostic feedback. The deployment leverages open-source frameworks, balancing resource availability and skillset requirements to benefit both developers and end-users. Our project aims to revolutionize cacao farming through precise, stable, and ethical deep learning solutions, ultimately enhancing crop resilience, productivity, and the livelihood of Ghanaian farmers.