Classification and Health Prediction in Plants Using Deep Convolutional Neural Networks

Abstract

Plants are one of the core components of human life and its surrounding environment. Various diseases not only diminish the ecological relevance of plants and the products they produce but also have an impact on their economic value. The main goal of this study is to identify plant kinds and design a suitable and effective method for judging a plant's healthiness based on pictures of its leaves in order to give a workable system for an instant and economical solution to this problem. The analysis of both biotic and abiotic elements that affect a plant's general health is known as plant pathology. Farmers must identify the issue quickly in order to take appropriate measures and stop additional losses. Consequently, it is recommended for this study to use a Deep Convolutional Neural Network (DCNN) to categorise damaged leaves. A genuine dataset of 4503 images of the 12 diverse tree leaves that were gathered at the “Shri Mata Vaishno Devi University in Katra, J&K, India”, is used to verify this work. Both healthy and unhealthy leaf photos are taken in the dataset that belongs to the 12 different plants type. The result found after testing the model was quite good. The suggested DCNN model will have greater classification accuracy and able to classify the plant type of the leaves as well as the model can also predict whether the leaf is healthy or unhealthy.