The identification of medicinal plants is crucial for traditional medicine and botanical research, yet it often requires significant expertise and time. This study leverages advanced deep learning techniques to automate and enhance the classification of 40 distinct medicinal plant species. We explore the efficacy of Convolutional Neural Networks (CNN), MobileNet, and a hybrid model combining MobileNet with Recurrent Neural Networks (RNN) for this task. Our approach involves training these models on a diverse set of plant images and evaluating their performance in terms of accuracy, precision, and recall. The CNN provides a robust baseline for image classification, while MobileNet offers an optimized solution for resource-constrained environments. The hybrid MobileNet+RNN model is investigated for potential advantages in sequential or contextual feature extraction. The findings aim to advance automated plant identification systems, making them more accessible for researchers, herbalists, and practitioners, ultimately improving the efficiency and reliability of medicinal plant classification.

Keywords: Medicinal plants, deep learning, Convolutional Neural Networks (CNN), MobileNet, hybrid model, image classification, medicinal plant identification.