Identification and Solutions for Grape Leaf Disease Using Convolutional Neural Network (CNN)

Abstract

Crops are facing many diseases nowadays. These diseases lead to major damage and economic loss and hence early detection of disease is necessary to prevent the damage acquired by the crops. Fallacious diagnosis and severity of diseases lead to improper use of pesticides. Expert naked eye observation is the primary method used by plant disease detection and identification. Yet an examination of the naked eye is time-consuming, costly and takes a lot of effort. Classification is a system by which the leaf is classified according to its unique morphological characteristics. There are so many methods of classification, choosing a classification system is always a daunting task, because the consistency of the result differs according to different input data. Image processing is one of the commonly used methods for the identification and diagnosis of plant-leaf diseases. More effective methods are developed and proposed for the early detection of plant disease with the lowest processing period due to technological and scientific advances. The main aim of the proposed system is to detect the disease of the grape leaf using a neural network algorithm and to provide preventive or solutions to the user. Earlier one had to manually extract the features from images and pass it to different classification algorithm but CNN performs both image recognition and feature extraction. Compared to other algorithms the accuracy for image classification by CNN is more.