Leaf Disease Detection Using Machine Learning (ML)

Leaf Disease Detection Using Machine Learning (ML)

C.V. Suresh Babu, Ambati Swapna, Dama Swathi Chowdary, Burri Sujit Vardhan, Mohd Imran
Copyright: © 2023 |Pages: 12
DOI: 10.4018/978-1-6684-9231-4.ch010
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

This method's central idea is the generation of features using grey level co-occurrence matrices (GLCM). The spatial interactions between pixels are to be measured by the matrices. A grey-level co-occurrence matrix is used to extract co-occurrence features. Texture classification can be used for a number of applications, such as pattern identification, object tracking, and shape recognition, when done correctly and accurately. The images of the leaves are used to identify plant diseases. As a result, it is beneficial to apply image processing techniques to identify and categorise illnesses in agricultural applications. Making predictions or judgements without being explicitly programmed is possible by utilising machine learning algorithms to create a model based on test data, also referred to as “training data.” Because it is very challenging for humans to detect disease in leaves, we have introduced a classification of plant leaf diseases in this project. This study extracts the leaf's textural properties and compares them to classifiers that use RF, LDA, NN, and CNN.
Chapter Preview
Top

Review Of Literature

Define the Problem

In some sections of a plant, such as the leaves, the symptoms of plant diseases are obvious. It is laborious to manually diagnose plant illness using photographs of the leaves. Therefore, it is necessary to create computational techniques that will automate the disease detection and classification process using leaf images.

Problem Statement

One of India's main economic sectors is agriculture. More over 50% of the workforce in India works in the agriculture industry. India is thought to be the world's top producer of pulses, rice, wheat, spices, and goods derived from spices. How well farmers run their companies is based on the calibre of the products they produce, which is reliant on plant growth and production. Consequently, identifying plant disease is crucial in the agricultural industry. The environment of the farmer is impacted by plant diseases since they might impede a plant's growth. It is advantageous to use automatic disease detection methods to identify plant diseases in their early stages symptoms of disease in plants (Gaidhane et al., 2018).

Background Work

Despite the difficulties listed in the problem statement, research on plant disease detection is still ongoing. Over the years, a lot of different strategies have been put forth. Using Support Vector Machine techniques, plant diseases can be detected and differentiated in traditional systems. This method was used to diagnose diseases in sugar beets, and the classification accuracy ranged from 65% to 90%, depending on the disease kind and stage. K-means was employed as a clustering algorithm together with another approach based on leaf pictures and using ANNs as a technique for an autonomous detection and classification of plant diseases. There were 10 secret layers in the ANN. Six outputs were produced, with each representing a class that included five disorders in addition to the case.

Materials and Techniques

The framework that will be used in this work is shown in Figure 1.

Data Collection and Dataset Preparation: When searching online for images, use the keywords “plant names” and “disease names.” All of the photographs can then be categorised into several groups.

Complete Chapter List

Search this Book:
Reset