Disease Detection in Plant Leaves using K-Means Clustering and Neural Network

International Journal of Trend in Scientific Research and Development (IJTSRD)
Volume 4 Issue 1, December 2019 Available Online: www.ijtsrd.com e-ISSN: 2456 – 6470
@ IJTSRD | Unique Paper ID – IJTSRD29562 | Volume – 4 | Issue – 1 | November-December 2019 Page 442
Disease Detection in Plant Leaves using
K-Means Clustering and Neural Network
P. Harini1, V. Chandran2
1B.E, Student, 2Assistant professor
1,2Department of Electronics and Engineering, Bannari Amman Institute of Technology,
Sathyamangalam, Coimbatore, Tamil Nadu, India
ABSTRACT
The most contributing variable for the Indian Economy is Agriculture yet at
the same time there is absence of mechanical improvementinmany partsofit.
The harm caused by rising, re-developing and endemic pathogens, is vital in
plant frameworks and prompts potential misfortune. The harvest generation
misfortunes its quality because of much infections and some of the time they
happen however are indeed, even not obvious with stripped eyes. Plant
malady recognition is one such dull process that is hard to be inspected by
exposed eye. This paper shows an answer utilizing image processing
calculations by loading the image, preprocessing and feature extraction using
K-means clustering and segmentation method to identify the disease with
which the plant leaf been affected.
KEYWORDS: Image processing, pre-processing, feature extraction, K-means
clustering, Segmentation
How to cite this paper: P. Harini | V.
Chandran "Disease Detection in Plant
Leaves using K-Means Clustering and
Neural Network"
Published in
International Journal
of Trend in Scientific
Research and
Development(ijtsrd),
ISSN: 2456-6470,
Volume-4 | Issue-1,
December 2019, pp.442-445, URL:
https://www.ijtsrd.com/papers/ijtsrd29
562.pdf
Copyright © 2019 by author(s) and
International Journal ofTrendinScientific
Research and Development Journal. This
is an Open Access article distributed
under the terms of
the Creative
CommonsAttribution
License (CC BY 4.0)
(http://creativecommons.org/licenses/by
/4.0)
I. INRODUCTION
The foundation of Indian economy is farming. Right around
78% individuals depend on farming. Horticulture
contributes real offer in the development of Gross
Development product (GDP) of the nation. Trim sicknesses
contribute specifically and in a roundabout way to the
spread of human irresistible illnesses and natural harm. As
these infections are spreading and making harm, the typical
working of the plant is affected. In the field of horticulture,
plant sickness identification assumes an imperative job in a
legitimate way. General checking of the plant that is
developed is prime undertaking for eachrancher.Keeping in
mind the soundness of plants assumes a critical job. The
traditional strategy utilized is the discovery of malady
utilizing bare eye, in this technique an extensive number of
specialists are required though in image processing, this
confinement has been survived.
This paper gives an unmistakable view about the job of
image processing in the recognition of plant infection In this
phase, k-means clustering algorithm is used to detect the
disease of the plant. The image of the plant is loaded, that is
needed to be checked. The imageisthensegmentedfollowed
by the cluster details. The above datas are processed with
iterations. The name of the disease that attacked the plant
along with the data for accuracy will be shown.
II. DISEASE DESCRIPTION
1. Althernia Alternata:
It is a fungal disease. It causes leaf spots, it affects more than
300 species of plant. It needs a warm and humid
atmosphere. The symptom is browning and yellowing of
leaves.
Fig 1: a leaf affected by althernia alternate
IJTSRD29562

International Journal of Trend in Scientific Research and Development (IJTSRD) @ www.ijtsrd.com eISSN: 2456-6470
2. Anthracnose:
It is a fungal disease, it causes withering of tissues. The
severity varies from mild attack to death. Trees like oak,
maple are affected.
Fig2: Anthracnose disease in a leaf
3. Bacterial Blight:
It is caused by bacterial pathogen. The bacteria enter
through the stomata or any wounds. The early symptom is
the brown spots on the margin. Then the centre of the leaf
will turn dark reddish brown and dry out.
Fig3: leaf with Bacterial blight attack
4. Cercospora leaf spot:
It affects the smooth texture of the leaf. First the spot occurs
on the leaf at the bottom of the plant and after thatitspreads
upward. Cercospora is less severe under shady conditions.
Fig4: Cercospora spotted in a leaf
III. LITERATURE SURVEY
1. The proposed technique involves development of the
new spectral indices for identifying the winter wheat
disease. They consider three different pests (Powdery
mildew, yellow rust and aphids) in winter wheat for
their study. The most andthe leastrelevant wavelengths
for different diseases were extracted using RELIEF - F
algorithm.
2. The proposed technique uses image processing for
detection of disease and the fruit grading. It includes
artificial neural network for detection of disease.
3. The proposed technique includes capturing of the chilli
plant leaf image and processed to determine the health
status of the chilli plant.
4. The proposed technique includes detection of plant
disease using otsu algorithm and k-means clustering
algorithms.
5. The proposed system used FPGA and DSP based system
developed by Chunxia Zhang, Xiuqing Wangand Xudong
Li, for monitoring and control of plant diseases.
IV. METHODOLOGY
Fig 5: Proposed workflow
1. IMAGE ACQUISITION:
The procedure includes coordinate obtaining of pictures
from any equipment sources orfromanydatabasewhich has
been as of now nourished. This is the initial phase in the
work process of picture handling. This progression of
gaining any preforming picture is preeminentvital inlight of
the fact that with no picture none of the accompanying
procedure + should be possible. This picture is totally a
natural picture.
2. IMAGE PREPROCESSING:
The main motive of going for image pre-processing is to
improve the image data .It is used to supress unwanted
distortions .It also enhances the image features
Pixel brightness transformations
Geometric transformations
Grey scale transformations
3. IMAGE SEGMENTATION:
It is the process of dividing an image into multiple parts.The
main aim of segmentation is to simplify the image into a
form that is much easier to analyse. Different types of
segmentation are
IMAGE ACQUISTION
PRE-PROCESSING
SEGMENTATION
FEATURE EXTRACTION
DETECTION &
CLASSIFICATION OF PLANT
DISEASE

Region based segmentation
Threshold segmentation
Edge detection segmentation
Regional growth segmentation
4. FEATURE EXTRACTION:
The features of the image include colour, texture,
morphology, edges etc. Initially some data is measured and
by using it some derived values are obtained thatgivessome
information about the image.
5. DETECTION AND CLASSIFICATION:
After feature extraction classification is
done by using neural network. The output of the neuron is
the weighted sum of the inputs. Once the neural network
weights are fixed, it can be used to compute output values
for new images.
3
Fig6: Back propagation network
K-MEANS CLUSTERING:
It is used to classify the object into K number of classes.
Algorithm:
Select the centre of K cluster
Minimise the distance between the pixel and cluster
centre by assigning every pixel in the image
Average all the pixels in the cluster and compute the
cluster centre.
Repeat this process until convergence is obtained.
Fig7: Cluster selection and display
V. RESULT AND ANALYSIS
Plant disease detection and identification using image
processing is a major breakthroughtosolvemanychallenges
of agriculture by the farmers. The method evolved with this
paper came up with a greater accuracy.
STEP 1:
First the query image is loaded by the user.
STEP 2:
The brightness and other attributes of the image is
enhanced.
STEP 3:
In this final step, the image is segmented and the cluster
number is chosen followed by the classificationoftheimage.
The result is displayed along with the affected region and
accuracy.
Result Analysis for Althernia Alternata:
The accuracy analysis and plot ofAlternaria Alternata isvery
unique and it is highly efficient. The average accuracy is
calculated by considering ‘n’ number of leaves and the
accuracy plot is drawn below.
Disease name Leaf number Accuracy
Althernia
Alternata
1 98.5
2 97.3
5 96.6
4 96.6
5 95.8
Average 95.8
Table1: Accuracy table for althernia

Fig8: Accuracy plot for althernia alternate
VI. FUTURE WORK
The precise recognition and grouping of the plant infection
are vital for the effective development of plants and this is
possible by utilizing image processing. This paper examines
different procedures to section the infected part oftheplant.
It also talked about some feature extraction and grouping
procedures to separate the tainted leaf and the
characterization of plant ailments. The utilization of ANN
techniques for order of sicknessinplantsisdone.Fromthese
strategies, we can precisely recognize and arrange different
plant ailments utilizing image processing procedures.
In future we have planned to increase the efficiency and
accuracy of the disease detection by using other algorithms
like Otsu algorithm. Also, we will increase the number of
diseases that could be identified by this algorithm.
VII. REFERENCES
[1] Wenjiang Huang, Qingsong Guan, Juhua Luo, Jingcheng
Zhang, Linsheng Huang, Dongyan Zhang, "New
Optimized Spectral Indices for Identifying and
Monitoring Winter Wheat Diseases", IEEE journal of
selected topics in applied earth observation and
remote sensing, IEEE vol. 7, no. 6, June 2014.
[2] Dr. K. Thangadurai, K. Padamavathi, "Computer Vision
image Enhancement For Plant Leaves Disease
Detection", World congress on Computing and
Communication Technologies, IEEE 2014.
[3] Monika jhuria, Ashwani Kumar, Rushikesh Borse,
"Image Processing For Smart Farming: Detection Of
Disease And Fruit Grading", Second International
Conference on image information processing, IEEE
2013.
[4] Mrunalini R. Badnakhe, Prashant R. Deshmukh,
"Infected Leaf Analysis and Comparison by Otsu
Threshold and k-Means Clustering", International
Journal of Advanced ResearchinComputerScienceand
Software Engineering, vol. 2, no. 3, March 2012.
[5] H. A1-Hiary, S. Bani-Ahmad, M. Reyalat, M. Braik, Z.
Alrahamneh, "Fast and Accurate Detection and
Classification of Plant Diseases", International Journal
of Computer Applications, vol. 17, no. 1, March 2011.

Disease Detection in Plant Leaves using K-Means Clustering and Neural Network

More Related Content

What's hot (20)

Similar to Disease Detection in Plant Leaves using K-Means Clustering and Neural Network (20)

More from ijtsrd (20)

Recently uploaded (20)

Disease Detection in Plant Leaves using K-Means Clustering and Neural Network