Plant Detection and Classification Using Fast Region-Based Convolution Neural Networks

Abstract

The identification of plants is an important task for preserving plants, which are very crucial for the existence of humans on earth. Many of the species are at the stage of extinction and are present in very discreet location. Their identification and protection are one of the major concerns. Many of these plants have great medicinal value, so saving them becomes very important. Plants can be identified using their leaves, bark, seed, fruit, flower, etc. The methodology described in this paper considers the identification of plants using the features of its leaves. The plants considered are the medicinal plants which can be presented in discreet locations like the Himalayas or can be presented in the kitchen garden. In this paper, we have used regional convolution neural network (RCNN) for the identification of plants. The system uses fast RCNN (fast RCNN) model using convolution networks to extract features and for classification support vector machine (SVM) issued.