Identification of Enhanced Green Natural Environmental Plant Species Using Deep Learning

Abstract

Many individuals have been trying to resolve the issue of recognizing weeds for several years. They used a broad variety of methods to identify weeds from ordinary crops with a common goal, but no system has been created that has made a business breakthrough. Botanists and those who study plants however, are able to identify the type of plants at a glance by using the characteristics of leaves but we are interested in identification of plants by using the digital cameras, mobile devices, like techniques of image processing and the pattern recognition. Features are shape; colour and texture have to be studies for the plant identification. This work proposes a good computer vision deep learning technique of convolution neural network consists of four layers, convolution layer, max pooling, dropout, average pooling. The optimum results were accomplished with much less computational effort and it shows the efficiency of the algorithm. We implemented the proposed model by using the Kaggle tool. We discussed the result and the accuracy of our work. We can come out with the 86% accuracy for the plant identification using plant image technique. This paper addresses the technique that is used for the identifying plant using plant leaves. This paper discusses the model for plant identification.