Cacao Ripeness Detection System
Software as a Service
NoCategory
Thesis DesignClients
Southern Luzon State University
Techstack
PythonRaspberry Piscikit-learnpandas
Purpose
Southern Luzon State University needed a way to judge cacao pod ripeness objectively rather than by eye, so farmers can time harvests for better yield quality and consistency.
Description
A Raspberry Pi prototype that classifies cacao ripeness with a Naive Bayes model built in scikit-learn. The team assembled measurement datasets of pod characteristics, trained and evaluated the classifier with accuracy, precision, recall, and F1 scoring, then deployed the saved model to the Raspberry Pi where it classifies new readings on the spot. Design-of-experiments notebooks document the model selection, and pandas handles the data pipeline end to end.
