Identification and Classification of Medicinal Plants Found in Rural Areas Through leaf Image and Their Benefit Analysis Using Deep Learning

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.