Deep Learning Models for Automatic Detection and Classification of Pathogens in Olive Leaves

The purpose of this paper is to develop a fast and an accurate object recognition models using Convolutional Neural Networks (CNN). These models are used in smart agriculture applications to automatically detect and classify plant pathogens, especially in olive leaves. We start by proposing a Deep Learning (DL) model based on the CNN architecture. Optimizations for model accuracy are then presented, overcoming the problem of overfitting with four different solutions. A comparison of the impact of these different solutions is also provided. The results that were obtained are satisfying, with an accuracy rate of 94%.