Detection and Classification of Olive Leaves Diseases Using Machine Learning Algorithms

Olive cultivation plays a crucial role in the social and economic life of Tunisia which represents around 15 percent of the total value of final agricultural production. Like other crops, olive trees are very vulnerable to several diseases. To fight against them, it is important to detect them at an early stage. To address this challenge, deep learning (DL) techniques applied to leaf photos seem to be useful tool as it was already used to diagnose diseases in many application fields. In this research, we develop a novel two-stage DL model to extract in the first stage the subtle olive leaves and identify in the second one discriminant features to recognize the disease type. Our approach shows high classification performances.