Open AccessClassification of Plants Leaf Diseases using Convolutional Neural Network
Author(s) -
Reem Mohammed Jasim Al-Akkam,
Mohammed Sahib Mahdi Altaei
Publication year - 2021
Publication title -
al-nahrain journal of science
Language(s) - English
Resource type - Journals
eISSN - 2663-5461
pISSN - 2663-5453
DOI - 10.22401/anjs.24.2.09
Subject(s) - convolutional neural network , computer science , artificial intelligence , deep learning , dropout (neural networks) , categorical variable , agriculture , machine learning , rgb color model , artificial neural network , function (biology) , production (economics) , convolution (computer science) , pattern recognition (psychology) , geography , macroeconomics , archaeology , evolutionary biology , economics , biology
Agriculture is one of the most important professions in many countries, including Iraq, as the Iraqi financial system depends on agricultural production and great attention should be paid to concerns about agricultural production. Because plants are exposed to many diseases and monitoring plant diseases with the help of specialists in the agricultural region can be very expensive. There is a need for a system capable of automatically detecting diseases. The aim of the research proposed is to create a model that classifies and predicts leaf diseases in plants. This model is based on a convolution network, which is a kind of deep learning. The dataset used in this study called (Plant Village) was downloaded from the kaggle website. The dataset contains 34,934 RGB images, and the deep CNN model can efficiently classify 15 different classes of healthy and diseased plants using the leaf images. The model used techniques to augment data and dropout. The Soft max output layer was used with the categorical cross-entropy loss function to apply the CNN model proposed with the Adam optimization technique. The results obtained by the proposed model were 97.42% in the training phase and 96.18% in the testing phase.