Classification of Medicinal Plants and Their Diseases: Application to the AI-MedLeafX Dataset

Abstract

Medicinal plants are essential to traditional medicine and affect more than 80% of the world's population. Their identification remains complex due to the diversity of species and diseases. This study presents an automatic classification model based on convolutional neural networks (CNNs) to distinguish 4 species of medicinal plants (Camphor, HariTaki, Neem, Sojina) and their health states (Bacterial Spot, Healthy Leaf, Shot Hole), i.e. 10 classes in total. The model was trained and evaluated on the AI-MedLeafX public dataset.

The model achieves an accuracy of 88.06% (11.94% error), with the main confusion between Camphor Shot Hole and HariTaki Bacterial Spot. These results confirm the feasibility of a

automated identification of medicinal plants.

In this paper we are trying to develop and evaluate a deep learning model based on convolutional neural networks (CNNs) for the automatic classification of medicinal plants and their diseases from leaf images.