Automatic Recognition of Madurese Herbal Medicine Rhizome Images Using the Efficientnet Convolutional Neural Network

Recognizing and classifying herbal medicine rhizome images are important in traditional medicine and pharmacology. Our study proposes an automated approach for recognizing Madurese herbal medicine rhizome imagery using the EfficientNet Convolutional Neural Network (CNN). The EfficientNet model is known for its optimal balance between accuracy and computational efficiency, making it a suitable choice for image recognition tasks. EfficientNet-B3 is utilized for automated recognition of Madura herbal rhizome image datasets, comprising five types: Temulawak (Curcuma xanthorrhiza), Kencur (Kaempferia galanga), Jahe (Zingiber officinale), Lengkuas (Alpinia galanga), and Kunyit (Curcuma xanthorrhiza) (Curcuma longa). The experimental results demonstrate the effectiveness of the proposed approach, showcasing its potential to assist herbal medicine practitioners and researchers in identifying and cataloging rhizome specimens with high precision. The automated recognition system for Madurese herbal rhizomes revolutionizes the selection process, enabling swift, consistent, and accurate identification during seeding and production. Our proposed system ensures seamless automation, enhancing the efficiency of herbal medicine practices. Moreover, our study will be essential in preserving and promoting traditional herbal medicine knowledge while optimizing production.