Classification and Assessment of Visual Content of Medicinal Plants Using CNN

Abstract

Use of medicinal plants has a long history in traditional medicine around the world. It is critical to correctly identify medicinal plants in order to determine their therapeutic characteristics and prospective applications. However, due to the complexity of their appearance, it might be a challenging process. Variation in features of leaf restrict the use of existing plants identification methods. This research suggests a deep learning-based solution using VGG-19 model, which employs a convolutional neural network (CNN) that is capable of learning and encoding complex aspects of images, allowing it to recognize and classify medicinal plants with high accuracy. The experiment has been done with the help of Flavia dataset which complies 1000 images, and extracted features are classified using different classifiers. The research has the potential to provide a knowledge base for identifying herbal plants to healthcare providers and herbal medicine researchers.