PLANT IDENTIFICATION AND CLASSIFICATION THROUGH CONVOLUTIONAL NEURAL NETWORKS: A DEEP LEARNING APPROACH

Abstract

Six commonly found native medicinal herbs from Tamil Nadu make up the dataset. For primary classification tasks, a powerful CNN (Convolutional Neural Network) model is used. Thirty percent of the dataset is set aside for validation and the remaining seventy percent is used to train the model. Techniques for dataset augmentation and randomization are used before model input. Using SVM (Support Vector Machine) classifiers in conjunction with a one-versus-all coding architecture, the ECOC (Error Correcting Output Codes) framework is put into practice. CNN models are used to extract features, which are then fed into the classification model.