Analysis and Classification of Medicinal Plants found in India

Ayurveda, the traditional medicinal system of India, suggests using various plant parts such as leaves, roots, fruits, and stems to cure a variety of ailments. Indian plants and herbs have been studied for their potential as medicines by numerous academics in the west. Due to this the recognition of these plants has become a focus of ongoing research. For the goal of classification and identification, this research employs an appropriate machine learning approach. Several machine learning approaches have been employed for this purpose over the year. The three phases of the suggested methodology are preprocessing, feature extraction, and classification. The preprocessing phase includes standard image processing operations including gray scale conversion and border improvement. In feature extraction the phase gaps were closed using Closing operation of Morphological Transformation and the extraction of 15 features that were used for classification. In this study we aim to use Random forest algorithm for classification of 29 plant species which are naturally occurring and readily available in India for their medicinal purpose. During the course of this paper we chose to select the most accurate model, by conducting a comparative study for classification by implementing SVM, KNN, Gaussian Naive Bayes and Random Forest Classifier algorithms which gave 85%, 73%, 85% and 87% accuracy respectively.