Tomato Plant Leaf Disease Classification based on CNN features and Support Vector Machines
Amine Mezenner, Hassiba Nemmour, Youcef Chibani, Adel Hafiane
Abstract
Early identification of plant leaf diseases constitutes an effective way for protecting and improving the crops production. Recently, there was a growing interest in developing intelligent systems that achieve plant disease detection with high performance. Various popular models of machine learning techniques as well as Convolutional Neural Networks (CNN) were successfully used on various datasets. The present work aims to develop a robust Tomato disease classification system that combines CNN features with Support Vector Machines (SVM). Experimental analysis is conducted on a public dataset of leaf images representing healthy tomato and nine tomato diseases. The results obtained reveal that the proposed system achieves similar and commonly higher performance than several state of the art systems.