OPTIMIZING APPLE ORCHARD MANAGEMENT WITH MACHINE LEARNING: DIAGNOSTIC MODELS FOR HARVESTING

Abstract

Efficient management of apple orchards is crucial for ensuring optimal fruit yield and quality. However, managing apple orchards has various factors such as pest control, extreme weather effects, and predicting the best harvest time, which remains a challenge for orchard managers. This thesis aims to apply machine learning techniques to optimize orchard management, focusing on diagnostic models that analyze the intrinsic state of individual trees. By using images of the trees, the model can separately identify the condition of each main part of the apple tree (apples and leaves) allowing it to generate a comprehensive description of the tree’s overall health. This model can potentially reduce costs, increase productivity, and promote sustainable orchard practices. The research demonstrates how machine learning can be used in agriculture to assess the individual state of trees in orchards, allowing managers more precise control over orchard maintenance.