Deep Learning-Based Plant Disease Detection Using Android Apps

Deep Learning-Based Plant Disease Detection Using Android Apps

Mohd Shoaib, Firoz Ahmad, Mohd Ammad Rehman
Copyright: © 2022 |Pages: 21
DOI: 10.4018/978-1-6684-5141-0.ch009
OnDemand:
(Individual Chapters)
Available
$33.75
List Price: $37.50
Current Special Offers
10% Discount:-$3.75
TOTAL SAVINGS: $3.75

Abstract

Disease detection in plants is crucial for preventing losses in yield and agricultural productivity. Historically, disease identification has been supported by agriculture extension organizations, which were difficult to access from villages. Farmers have to go to their field and manually monitor plant disease. The aim of this work is to develop an Android application that provides an easy-to-use platform for farmers to identify diseases in their crops. The mobile application will help to take responsive action according to the disease detected in their plants and can be easily used by anyone who is interested in analyzing the disease of the plants. This work reports on the classification of 26 diseases in 14 crop species using 54,306 images from PlantVillage dataset using a convolutional neural network approach. The models used are Inception-v3 and MobileNet. The correct prediction of the correct crop-diseases pair in 38 classes decides the criteria for performance measurement. The most accurate model achieves an overall accuracy of 96.32%.
Chapter Preview
Top

Literature Review

This section briefly discusses the various machine learning techniques, and the various approaches used in plant disease detection. Recognizing the combination of smart phones and advances in computer vision, using convolutional neural network together to identify 14 crop species and 26 diseases (Mohanty et al. 2016). The trained model achieves an accuracy of 99.35%. Another approach of plant disease recognition model, developed by Berkley Vision and Learning Centre based on leaf image classification using deep learning framework called Caffe (Sladojevic, 2016; Nooraiyeen, 2020). ImageNet, a large-scale ontology of images database is used in three easy applications of object recognition, image classification and automatic object clustering (Deng, 2009).

Complete Chapter List

Search this Book:
Reset