Coffee rust detection and recommendations module through deep learning algorithms

Abstract

In recent years, coffee has faced a global increase in plagues and diseases, impacting the quality and profitability of its benefits. One of the most devastating and common diseases is coffee rust, which has caused a significant decrease in coffee production at the national and state levels by causing the falling of mature and young leaves, reducing production. In this work, we present a mobile application development using a deep learning algorithm model to identify and predict the level of coffee rust infection, providing treatment and prevention recommendations for coffee farmers. The model classified with an accuracy of 58% of rust diseased leaves in a dataset of diseased and healthy leaves.