ARCHIVES

Medicinal plants play an important role in traditional healthcare, but identifying plant species from their visual appearance can be challenging for people without specialized botanical knowledge. Manual identification may require considerable time and access to expert references. This paper presents PlantRx, an AI-based medicinal plant recognition system designed to identify Indian medicinal plants from uploaded images and provide relevant herbal information. The system uses the EfficientNet-B0 deep learning model to classify images into 40 medicinal plant classes. After classification, it displays the predicted plant name and confidence score, along with information about medicinal properties, traditional uses, plant parts, preparation methods, benefits, and safety information. A web-based interface developed using Python and Flask provides access to the system, while MySQL supports the storage of user information, plant details, and prediction history. PlantRx aims to improve access to medicinal plant identification and educational information through a centralized platform. Quantitative model accuracy and comprehensive testing results must be established using verified experimental data before the system's performance can be conclusively assessed.