IDENTIFICATION OF AYURVEDIC MEDICINAL LEAVES USING DEEP LEARNING

Ayurvedic medicine is an ancient kind of medicine. In the last few years, it has regained prominence. Ayurvedic botanicals are used to make medicines used in this therapeutic approach. These plants come in a wide range of types and must be distinguished from many other plant varieties found in nature. It is exceedingly difficult for an ordinary man to identify locally available medicinal herbs without sufficient knowledge. This presentation will present a new technique for recognizing Ayurvedic medicinal plants leaf using convolutional networks (CNN) and photos of leaves. Computer vision and the computer technology that supports it has advanced to the point that it may now be employed in a variety of fields. Image classification is one of its uses, which recognizes images more precisely than traditional methods. All the necessary information and steps used throughout the implementation process are described in the paper. All the essential steps like gathering images to create a database, training the models are described in detail. The deep neural network method that we used gives a more accurate classification than other classification methods. The advantage also lies in its simplicity, there is no need of preprocessing the images to extract features and then provide them to the model. A deep convolutional neural network can indeed be fed raw images as input. We don't need to extract visual attributes because neurons extract and store them as images transit through multiple levels in a deep neural network, allowing for accurate leaf classification. As a result, leaves are sorted and displayed via deep learning and a web app. The Deep learning technique that is being used in this paper is a Convolutional neural network.