Identification of Medicinal Plants Using Image Processing and Machine Learning

Plant recognition is both an indispensable and significant autecological method to botany taxonomy, which would combine recent advancements in Machine Learning with botanical science. In this study, medicinal plant identification was studied as primary because it plays a major role in the pharmaceutical sector. We also study other machine-learning approaches besides image processing methods to extract leaf features. The implemented system aims to provide timely, precise classification of medicinal plants. Indian leaf species were identified through Convolutional Neural Networks (CNNs) as a NET framework. The Guided Active Contour method also allows for simplification of leaf segmentation. The results from the trials showed accuracy similar to when traditional machine learning methods are used. This study would offer a feasible way to decrease human labor in pharmacognosy research and the botanical industry, enhance productivity, and automatically identify medicinal plants.