PLANT DIEASE DETECTION TECHNIQUE

Contents

• Seminar Motivation
• Problem Statement
• Introduction
• Aim and objective
• Literature Survey
• Working
• Advantages & Disadvantages
• Conclusions
• References
Indira College of Engineering Management, Pune 1
 Seminar Details
 Seminar Title : Plant Disease Detection Techniques
 Seminar Domain: Diagnosing dieases in an accurate and timely
way is of utmost important.

Indira College of Engineering Management, Pune

 Motivation
 One of the important sectors of Indian
Economy is Agriculture. Employment to
almost 50% of the countries workforce is
provided by Indian agriculture sector. India is
known to be the world's largest producer of
pulses, rice, wheat, spices and spice products.
Farmer's economic growth depends on the
quality of the products that they produce,
which relies on the plant's growth and the yield
they get.
 As agriculture is the largest occupation of our
country these techniques will be very helpful
for our farmers. India will be having all
technical methods in agriculture field which
will reduce farmers work into smart working.

Indira College of Engineering Management, Pune

 Problem
statement
 Important agriculture crops are threatened by a wide variety of plant diease and pests.

 about 42% of worlds agricultural crops are destroyed by yearly diease and pest .

 traditonal method of identifying plant pathogens is through visual examination.

 this is possible only if major damage has already been done to the crop, so treatments will be
limited or of no use.

 to save crops from irrepairable damage by pathogens farmers have to be able to identify the
infection even before it becomes visible.

 Hence, it is required to develop computational methods which will make the process of
disease detection and classification using leaf images automatic.

Indira College of Engineering Management, Pune

 Introduction
Major role in economic development of India Agriculture is the largest economic sector. The manual
classification and identification methods which are being used to distinguish between different types of leaf
diseases that are trusting on human resource. They are subjected to some kind of errors since these techniques
are focused by human involvement. Since humans are subjected to tiredness and the automated system also helps
to reduce the time consumed by manual techniques.

The deficiency of labors, automatic system needs to be

incorporated to minimize the work and Many new farming
computerization tools are being established by university
investigators that pose questions cabout the Effectiveness
with which we succeed current farming practices.

Indira College of Engineering Management, Pune 2

 Aim and Objectives of the Seminar
 The objective is to review various techniques of plant disease detection and discuss in terms of
various parameters. It is organized into the following sections. First section gives a brief introduction
to the importance of plant disease detection. Second section discusses the existing work carried out
recently in this area and also reviews the techniques used. Section three includes basic methodology
followed for developing disease detection system.
 Diagnosis is one of the most important aspects of a plant disease. Without
proper identification of the disease and the disease-causing agent, disease control
measures can be a waste of time and money and can lead to further plant losses.
Proper disease diagnosis is therefore vital.
 diagnosis of plant 
 disease control 
 Proper disease diagnosis 

Indira College of Engineering Management, Pune

We are using image processing techniques to classify
diseases & quickly diagnosis can be carried out as per
disease.
The basic methodlogy

Indira College of Engineering Management

 Literature
Survey
 First survey presents classification and detection techniques that can be used for plant leaf disease
classification. Here preprocess is done before feature extraction. RGB images are converted into
white and then converted into grey level image to extract the image of vein from each leaf. Then
basic Morphological functions are applied on the image. Then the image is converted into binary
image. After that if binary pixel value is 0 its converted to corresponding RGB image value. Finally
by using Pearson correlation and Dominating feature set and Naïve Bayesian classifier disease is
detected.
 This section reviewed the different work carried out in area of leaf disease detection. Many
researchers worked in the field of implementing computer enabled diagnosis of leaf disease in
different plant. The following table 1 represents the type of leaves, algorithm used for classification,
accuracy reported and the type of disease detected.
 In recent years, many researchers have been applied computer vision and machine learning for
detecting plant leaf disease. Many of them considered these techniques as an effective architecture as
it provides feature extraction without manual intervention

Indira College of Engineering Management, Pune

 Literature Review for Plant Leaf Disease Detection and
Classification
Author and Year of Types of Algorithm Used for Reported Types of disease
Publication leaves Classification Accuracy detected
Mukherjee et al, Tulsi Leaves Multi-Layer Back 80% Health of tulsi leaves
IEEE Paper 2019 PropagationPerceptron
Renugambal & Sugarcane leaf SVM 91% Sugarcane mosaic
Senthilraja, IEEE virus, Brown spot,
Paper 2019 Red rot, Leaf spot,
Brown spot
Jothiaruna, et al., General Plant Region Growing and Color feature 87% Disease affected
IEEE Paper 2019 disease images Extraction portion
segmentation
G. Saradhambal, A general plant Otsu’s algorithm assumes image to 95% The shape oriented features
used
et.al IEEE Paper contain two classes of pixels i.e. it werearea, color axis length,
2018 forms bimodal histogram with the eccentricity,
solidity and perimeter whereas,
foreground and background pixels. contrast,
For feature extraction, shape and correlation, energy,
homogeneity and
texture oriented features were mean were the texture oriented
used. features.

Indira College of Engineering Management,

Pune
System Architecture
Block Diagram

Indira College of Engineering Management

 WORKING
A. Image Acquisition:
In Acquisition Process Diseases images of the plants are capture through the High Resolution camera.
This image is in RGB(Red, Green and Blue) form. Color conversion structure for the RGB leaf image is
created, and then, a device independent color space conversion for the color variation manufacture is
applied such as HIS model .
B. Image Pre-processing:
To remove noise in image or other object removal, Image clipping i.e. cropping of the leaf image to get
the interested image region. Image smoothing is done using the smoothing filter. Image enhancement is
carried out for increasing the Contrast.A. Image Processing Method Principle: In image processing
method leaf area is calculated through pixel number statistic. Unit pixel in the same digital images
represent the same size hence from known reference area and pixel count, unit pixel size can calculate,
so that it is easy to calculate leaf area by counting total pixel in leaf area region.
C. Image Segmentation:
Segmentation means partition of image into diverse part of same skin tone or having some likeness
dissection means parceling of picture into different part of same elements or having some likeness. The
division should be possible utilizing different.

Indira College of Engineering Management,

Pune
 Continue..
D. Feature Extraction:
The input image is enhanced to protect information of the pretentious pixels before color from the
background. The color space equally is used to reduce effect of illumination and distinguish between
disease and non disease leaf color inventively the resulting color pixels are clustered to acquire groups of
colors in the image.
E. Classification:
In plant leaf categorization leaf is classified based on its different morphological facial exterior. Some of
the classification techniques used Component Analysis, k-Nearest national Classifier. Plant leaf infection
classification has wide application in cultivation

Indira College of Engineering Management,

Pune
Image segmentatiom
Fig. Basic Steps and Classification

Indira College of Engineering Management

 FLOW OF THE SYSTEM

1) Input Leaf Image

2) Pre-processing
 RGB Separation – we do AND operation with 0xff with each color value.
 Image Blurring - Remove Noise From image)
 Pre-processed Image
3) Apply Color Model Conversion (RGB TO HSV Conversion)
4) Histogram Plotting-(Obtain Frequency of color in image and map it in
buckets called histogram)
5) Comparing current features with Database store features (All the above
algorithms will be applied to
database images at the time of storing)
6) Apply distance Measures Equations (Cosine, Jaccard etc.)
7) Sort Score Vector (Bring top matching values at the top)
8) Classify Image as Infected and Non infected (Threshold Checking)
9) Show Output

Indira College of Engineering Management

Phases of plant disease detection system

Indira College of Engineering Management

 Classification techniques Comparison

Technique Advantages Disadvantages

K-Nearest Simpler classifier as exclusion of More training samples- more
Neighbor any training process. Applicable speed of computing distances
(KNN) in case of a small not trained sensitive to irrelevant inputs so
dataset. expensive testing every time.
Radial Basis Faster Training. It is slower in
Function Hidden layer is execution speed
(RBF) easier to interpret.

Support Simple geometric interpretation Slow training. Difficult to

Vector and a sparse solution. robust, understand. For classification
Machine when sample has some bias. large support vector required
(SVM)

Back Easy to implement. Learning can be slow.

propagation Applicable to wide range of It is hard to know how many
Network problems. Able to form arbitrarily neurons as well as layers are
(BPN) complex nonlinear mappings required.

Indira College of Engineering Management,

Pune
ADVANTAGES APPLICATIONS
 High accuracy
 Bio Farm
 Low complexity
 Bio-Pesticides
 Detection of image been classified
without any noise.
 Limitations

In order to obtain superior results in the detection of plant

disease, methods require a greater amount of data. This is
a drawback since currently available datasets are usually
small and do not contain enough images, which is a
necessity for high-quality decisions.

Indira College of Engineering Management,

Pune
 Conclusion

This presentation summarizes various techniques of disease detection using image processing that have
been used by a number of researchers in the past few years. The major techniques employed were:
BPNN, SVM, K-means clustering, Otsu’s algorithm, CCM and SGDM. These techniques are used to
detect if the leaves are healthy or diseased. Various challenges arise in this process including the
automation of the detection system using complex images captured in outdoor lightning and intense
environmental conditions. This review paper concludes that these disease detection techniques show a
efficiency and accuracy such that they have the ability to run the system developed for detection of leaf
diseases besides having some limitations. Therefore, there is a lot that can still be done in this field for
enhancement of the existing works.

Indira College of Engineering Management

 Reference
• M.Malathi , K.Aruli , S.Mohamed Nizar , A.Sagaya Selvaraj, “A Survey on Plant Leaf
DiseaseDetection Using Image Processing Techniques” International Research Journal of
Engineering and Technology (IRJET) e-ISSN: 2395-0056 p-ISSN: 2395-0072 Volume: 02
Issue: 09 | Dec-2015
• https://www.google.com/search?
q=aim+for+plant+disease+detection+techniques&sxsrf=ALeKk03cVTaqNx9RGlp47vGKd
NgElShUlQ
%3A1622740411532&ei=uw25YIeDIJb0rAG4mpDAAg&oq=aim+for+plant+disease+dete
ction+techniques&gs_lcp=Cgdnd3Mtd2l6EAM6BwgAEEcQsAM6BwgjELACECc6BggA
EAgQHjoFCAAQzQJQ1idY9TdgtjpoAXACeACAAf8BiAGlDpIBBTAuNC41mAEAoAE
BqgEHZ3dzLXdpesgBCMABAQ&sclient=gws-wiz&ved=0ahUKEwjHzKqO-
_vwAhUWOisKHTgNBCgQ4dUDCA4&uact=5
• Ms. Kiran, R. Gavhale, Prof. Ujwalla Gawande, “An Overview of the Research on Plant
Leaves Disease detection using Image Processing Technique” IOSR Journal of Computer
Engineering (IOSR-JCE) e-ISSN: 2278-0661, p- ISSN: 2278-8727Volume 16, Issue 1, PP
10-16, 2014. Indira College of Engineering Management
THANK YOU