Automated coffee plant detection and counting system using machine learning

This paper presents a machine learning-based system for automated coffee plant detection and counting using drone-captured images. Leveraging the YOLO (You Only Look Once) algorithm, the system aims to improve efficiency and accuracy in coffee farming, particularly in regions like Costa Rica, where manual plant counting is labor-intensive and prone to error. Our model achieves a mean precision of 87.03%, with a median precision of 87.28%. Notably, 45.45% of the results exceed 88% precision. The system's scalability allows it to be used across farms with varying drone equipment quality. The paper outlines the methodology, results, challenges faced, and future improvements aimed at extending the system's capabilities for other crops.