The decline in public knowledge about medicinal plants in Indonesia threatens cultural preservation and the utilization of biodiversity. This community service project aims to bridge this gap by developing a prototype of an automated plant identification tool. The community service process focused on designing and training an Artificial Intelligence-based system. The method used was deep learning with the Xception architecture, implementing transfer learning and fine-tuning strategies for optimization. All activities, from data collection to model testing, were carried out from September to December 2024. The result was a classification model with 98.05% accuracy. Consequently, this model has great potential to be developed into a publicly accessible application, supporting the preservation of traditional knowledge and botanical education.
Copyrights © 2025