Abstract:
A medicinal plant identification system based on the EfficientNetB0 transfer
learning model is suggested by the current study. People's trust in using medicinal
plants and their ability to correctly identify them have a big impact on their
utilization. Utilizing medicinal plants is crucial in the medical field to satisfy the
demand for medication. Inappropriate usage of medicinal herbs can lower
immunity and lead to a number of issues for humanity. To guarantee their effects,
it is crucial to properly identify and utilize our medicinal plants. In this paper, we
propose to use an automated deep-learning model to handle the problem of
manually monitoring plants to identify and classify medicinal plants, which is
initially quite difficult. In order to help people, recognize and utilize therapeutic
plants, this study aims to identify and categorize them. We gathered a clean and
high-quality dataset for the categorization and identification of medicinal plants.
Finding medicinal plants in the village vicinity and taking pictures of their leaves
was the first stage of the data collection process. We separated medicinal plant
species into ten distinct classes using several transfer learning models in order to
precisely identify and categorize medicinal plants. Nonetheless, the 99.87%
reliability score of the medicinal plant dataset demonstrates how highly effective
EfficientNetB0 is. Thus, this chapter's goal is to promote the use of medicinal
plants, boost their dependability, and guarantee their usage by correctly recognizing
them.
