Enhanced Medicinal Plant Leaf Classification via Transfer Learning: A VGG16-Driven Approach for Precise Feature Extraction

Medicinal plant identification is essential for holistic treatment systems, where accurate taxonomy ensures the safe and effective use of therapeutic properties. However, traditional methods often struggle with high feature variability and visual similarities among species, increasing the risk of misidentification. To address this, Herbplex, an Android application, employs the VGG16 deep learning model for transfer learning. VGG16's pretrained architecture extracts deep features, optimizing the learning process for accurate classification. The model was trained on a dataset of 35 medicinal plant species, achieving an impressive classification accuracy of 93%. Herbplex not only ensures precise taxonomy prediction but also provides detailed medicinal properties of identified plants, enhancing its utility in healthcare and agriculture. By facilitating reliable plant recognition, Herbplex supports advancements in natural medicine, botany, and therapeutic research. This innovative solution underscores the potential of AI in promoting sustainable plant based treatments, ensuring safety, and driving innovation in holistic healthcare systems.