You are on page 1of 2

used to extract the leaf disease feature.

After that Support Vector Machine (SVM) is


used to classify the leaf diseases.

1.2. Aim and Objectives


The aim of the system is to implement the plant leaf disease identification
system. To complete this system, the following objectives are performed.
 To identify the boundaries of the affected area.
 To create an automatic system for leaf disease detection by Image
Processing system.
 To extract the infected area by Image Segmentation.
 To classify the disease using a classifier.

1.3. Overview of the Thesis


For the real time application, the system requires the images of leaves
reacquired using high resolution camera so as to get the better results and efficiency.
This system consists of four main steps.
In this system, healthy and diseases images of leaf are put as the input to the
system. These input images are transformed into HSV color space and the color
feature is extracted from it. Only H component is taken for further analysis while
extracting the texture feature. As green part of the leaf is healthy, the green pixels are
masked and removed. After extracting the infected portion, it is segmented into a
number of blocks of equal size. Then texture feature analysis is done for the selected
segments and the resultant feature vector is given as input to the classifier.
In the feature extraction step, feature vectors such as color and texture are
extracted. In this system, color features are also extracted from Mean value and then
statistical texture features are obtained by using Gray level Co-occurrence Matrix
(GLCM). Texture features like Contrast, Energy, Homogeneity, and Correlation are
computed for the H-component.
Finally, Support Vector Machine (SVM) is used to identify the plant leaf
diseases. In this system, Multiclass Support Vector Machine (Multiclass SVM)
technique is used for classification and identification of leaf disease.

1.4. Outlines of the Thesis


This thesis is composed of five chapters. Chapter 1 includes introduction to
Plant leaf Diseases, aims and objectives, overviews of the thesis and outlines of the
thesis. Chapter 2 describes a brief literature review of leaf disease detection and the
previous researches of corresponding color transformation, segmentation, and feature
extraction and classification techniques. Chapter 3 highlights the theory background
of the leaf disease identification and classification which is applied for this research.
Chapter 4 represents the design and implementation of the system and experimental
results. Finally, conclusion and further extensions for this research are presented in
Chapter 5.

You might also like