International Conference on Communication and Signal Processing, April 4-6, 2019, India

Identification of Plant Disease using Image

Processing Technique
Abirami Devaraj, Karunya Rathan, Sarvepalli Jaahnavi and K Indira


Abstract—Agriculture has become far more than simply a In case of industrial sickness refers to waste of inventory
method to feed ever growing populations. It’s important disease is clearly shown as the traditional gries process
wherever in additional than seventieth population of an Asian physical performance of plantae kingdom. From the existing
country is depends on agriculture. Which means it feeds nice proof a symptom might be determined. Illness is caused by
range of individuals. The foremost necessary consider less
amount crop of quality because of disease. Detecting disease microorganism that is any agent inflicting illness. Illness
may be a key to stop agricultural losses. The aim of this project management may be a tough task[5-10].Largely the verities of
is to develop a software system answer that Mechanically find diseases are seen on the leaves on plant, fruits or stems of the
and classify disease. The step like loading an image, pre- plant. Due to the difficulty of visual patterns the exact se
Processing, Segmentation, extraction and classification are quantification of those visually diagnosed diseases, pests, and
involves illness detection. The leaves pictures are used for
traits has not studied. Diseases are detected on the leaves or
detecting the plant diseases. Therefore use of image process
technique to find and classify diseases in agricultural stems of the plant in most of the cases. Booming cultivation
applications is useful. of crops the recognition of plants, leaves and searching for
the diseases, symptoms of the illness affect, plays a necessary
Index Terms—Image Processing, Random Forest Classifier, role[11-16].To avoid a human interference developing a
plant disease, Segmentation, Extraction. computer vision system to find, recognize, and classify illness
affected on crops and therefore solution in proper unbiased
I. INTRODUCTION
call regarding illness infection and its more valuation.

A GRICULTURE is a backbone of our country. Farmers

Farmers to avoid consulting divine the event of an automatic
system helps. Automatic detection might prove gain in
have good selection of crops for his or her farm. Anyway, the looking huge fields of crops and leaf disease detection is most
crops cultivation for maximum profit and standard significant analysis topic and then from the symptoms that
manufacture is usually scientific. This might be developed by gift on the plant leaves, it can automatically notice the
the help of technical assist. The supervision that continually diseases[17-24].Image processing plays major Role.
recurring crops needs supreme power especially for the The MATLAB image processing starts with capturing of
disease management that may have a result on factors of digital high resolution pictures. Healthy and unhealthy
production significantly to make an economic profit. The pictures are captured and hold on for experiment. Then
process of image is the most excellent technique that having a pictures are applied for pre-processing for image
improvement. Captured leaf pictures are segmented using k-
paid job in agricultural application functions. Detect illness
means cluster technique to create clusters. Options are
from picture of plants[1-4].This effect can be mitigated with
extracted before applying K-means and Random Forest
the Aid for agricultural development Most of the primary Classifier for training and classification. Finally diseases are
symptoms are microscopic, so the identification disease of is recognized by this method. In this paper section one provides
restricted by human visual capabilities. This method is an introduction and importance of disease detection. Section
uninteresting, time intense. There’s need for style system that two provides a plant disease fundamental. Section three quick
mechanically acknowledges, classifies and quantitatively literature reviews of leaf illness detection techniques. Section
detects disease symptoms. four describes methodology of projected system based on
MATLAB image processing. Section five provides results and
discussion. Section six concludes this paper along with future
work.
The rest of this paper Section II describes the plant disease
Fundamental. Related work discussed in Section III. Proposed
work in Section IV. Result are discussed in Section V. In
Section VI gives antrocnose.finally concludes the paper in
Abirami Devaraj, Karunya Rathan, Sarvepalli Jaahnavi and K Indira are with Section VI. Cercospora Leaf Spot is discussed in Section VII
Information Technology, Sathyabama Institute of Science and Technology,
Chennai (e-mail: abiramidevaraj2512@ gmail.com,[email protected], and finally Concludes the Paper in Section VIII.
[email protected] and [email protected]).

978-1-5386-7595-3/19/$31.00 ©2019 IEEE 0749

 II. PLANT DISEASE FUNDAMENTAL Multi level Color Image Thresholding. Infection stage,
Generally, there are 3 forms of plant diseases. They’re spreading stage and worst stage are the result of RLB (Rice
microorganism, Viral and fungal in Fig. 1 Leaf Blast) shows in Fig. 2 illness is classed. The isn't
appropriate technique for detection of different diseases
which can have similar options.

Fig . 2. Rice Leaf Blast

This study [9] summarizes major image processing used for
identification of leaf diseases are k-means clustering, SVM.
This approach can considerably support an correct detection
of plant disease. There are 5 steps for the plant disease
identification that are same to be image acquisition, image
pre-processing, segmentation, feature extraction,
classification. By using this idea the illness identification is
completed for all types of leafs and additionally the user can
recognize the affected space of leaf in share by distinctive the
illness properly the user will rectify the matter terribly
straightforward and with less price.[11].The authors
introduced FCM Clustering Technique for Segmentation.
Fig. 1. Plant Diseases Fundamental SVM used for Classification of Plant diseases. Sickness level
of the accuracy with the help of this approach, it supply
III. RELATED WORK varied bar technique, the extent of destruction and helps to
Guiling Sun , Xinglong Jia , and TianyuGeng [6] discussed visualize whether or not the sickness spreads or not.[12]The
about plant disease detection using image recognition system authors introduced plant disease Detection in Image
supported multiple regression toward the mean is planned. processing mistreatment MATLAB. leaf & fruit pictures are
significantly, there square measure variety of innovations in segmental mistreatment k-means clustering methodology to
image segmentation and recognition system. In image make clusters. Options are extracted mistreatment GLCM
segmentation, an improved bar chart segmentation technique methodology. Apply K-means and SVM formula for coaching
that can calculate threshold mechanically and accurately is and prediction[24].Automatic and correct estimation of
planned. Meanwhile, the regional growth technique and true illness severity is important for food security, illness
color image processing square measure combined with this management, and yield loss prediction. Deep learning, the
technique to boost the accuracy and intelligence. While most recent breakthrough in computer vision, is promising
making the popularity system, multiple linear regression and for fine-grained illness severity classification, because the
image feature extraction square measure used. once methodology avoids the effortful feature engineering and
evaluating the results of various image training libraries, the threshold-based segmentation. Exploitation the apple plant
system is evidenced to own effective image recognition ability disease pictures within the Plant Village dataset that are
and prediction.M.N. Abu Bakar, A.H. Abdullah, N. Abdul additional annotated by botanists with four severity stages as
Rahim, H. Yazid, S.N. Misman and M.J. Masnan [7]. They ground truth, a continuous of deep convolution neural
described about Rice leaf Blast illness detection mistreatment networks are trained to diagnose the extremity of the illness.

0750
 IV. PROPOSED SYSTEM B. Image Preprocessing
The previous system will solely determine the kind of The main motive of image pre-processing is to enhance the
illness which affects the leaf. We are going to give a solution image information contained unwanted distortions or to
with less time and lead you throughout the project.The reinforce some image features for any processing. Pre-
architecture of the planned work is shown in below Fig. 3. processing technique uses various techniques like dynamic
The steadily planned method consists of leaf picture image size and form, filtering of noise, image conversion,
information assortment, pre-processing of these pictures; enhancing image and morphological operations. in this work,
segment of these pictures using k-means cluster methodology, we used numerous MATLAB code to change size of the
GLCM is used to extracting the feature and at last the random image, to enhance contrast and RGB to grayscale conversion
forest algorithmic used for coaching of method. as shown in Fig. 4 and Fig. 5 for additional operations like
making clusters in segmentation.

Image Acquisition

Image Pre-processing

Image Segmentation

Feature Extraction
Fig. 5. nContrasted image and HSI image

Classification C. Image Segmentation

Segmentation of image is that the technique for conversion
of digital picture into many segments and rendering of a
Fig. 3. Architecture Diagram
picture into something for easier analysis. using image
A. Image Acquisition segmentation is employed for locating the objects and
bounding line of that image. In segmentation, we tend to used
Loading of an image is that the 1st technique of digital
K-means cluster technique for partitioning of pictures into
picture process and it's represented as capturing the image
clusters during which a minimum of one part of cluster
through digital camera and stores it in digital media for
contain image with major space of unhealthy part. The k-
additional MATLAB operations. it's additionally an action of
means cluster algorithmic rule is applied to classify the
retrieving a picture from hardware, so it will be capable
objects into K variety of categories per set of features. The
further method. In our work, using camera we tend to
classification is completed by minimize the total sq. of
captured healthy and diseased pictures of leaf & fruit as
distances among information entities and therefore the
shown in Fig. 4. two for MATLAB image processing system.
particular cluster. Image is regenerate from RGB Color space
to L*a*b* Color space during which the L*a*b* area
contains of a luminousness layer 'L*', chromaticity 'a*' and
'b*'.
D. Classification
Classification technique is used to coaching and testing of
the leaf of the plants. The Random forest classifier is used for
classification.
V. RESULTS AND DISCUSSION
A. Alternaria Alternata
It is a fungus that impact the leaf spots over 380 has family
of plant. It will too impact leaf spots, rots, blight and
Fig. 4. Original Diseased leaf image alternative plant elements shows in Fig. 6.

0751
 .
Fig. 6. Alternaria Alternata Diseased leaf and Output

B. Bacterial Blight
Bacterial Blight shows in Fig. 7 is characterized by tiny,
pale inexperienced spots or streaks appeared as water-soaked. Fig. 8. Antarcnose Diseased Leaf and Output
The lesions can expand then seem as dry dead spots. it's
going to extend till the complete length of the leaf. VII. CERCOSPORA LEAF SPOT
Infested leaf has tan to brown spots with curled leaf margin
and afterward it withers. because the illness progresses, the
leaflets flip yellow and additionally with curled margins.
Spots on the leaf petioles, stems, and flower components
become elongated and have brown color. On heavily troubled
plant shows in Fig. 9 defoliation might occur.

Fig. 7. Bacterial Blight Diseased leaf and Output

VI. ANTRACNOSE
Anthracnose shows in Fig. 8 a bunch of fungal illness that
have an effect on a spread of plants in heat, humid areas.
Shade trees like sycamore, ash, oak, and maple are
particularly inclined, though the illness is found during a
variety of plants, as well as grasses and annuals. Fig. 9. Cercospora Leaf Spot Diseased Leaf and Output

0752
 VIII. CONCLUSION [11] S. Megha, R C. Niveditha, N. SowmyaShree, K. Vidhya “ Image
Processing System for Plant Disease Identification by Using FCM-
The present study deals with Alternaria Alternata, Clustering Technique” International Journal of Advance Research,
Antracnose, Bacterial Blight and Cercospora Leaf Spot these Ideas and Innovations in Technology.© 2017, www.IJARIIT.com
[12] Sandesh Raut , Amit Fulsunge “Plant Disease Detection in Image
automatic illness detection using image processing techniques Processing Using MATLAB”International Journal of Innovative
in MATLAB. It involves loading an image, image Research in Science, Engineering and Technology, Vol. 6, Issue 6,
preprocessing, image segmentation, feature extraction and June 2017 Website: www.ijirset.com
[13] Trimi Neha Tete, Sushma Kamlu “Plant Disease Detection Using
classification. Development of automatic detection system Different Algorithms” Proceedings of the Second International
using advanced technology like image process facilitate to Conference on Research in Intelligent and Computing in Engineering
support the farmers within the identification of diseases at an pp. 103–106, 2017.
[14] Rajneet Kaur et al ”A Brief Review on Plant Disease Detection using
early or initial stage and supply helpful data for its in Image Processing” International Journal of Computer Science and
management. we might prefer to extend our work additional Mobile Computing, Vol.6 Issue.2, February- 2017, pg. 101-106
on a lot of disease detection. www.ijcsmc.com
[15] VijaiSingh, A.K.Misra “Detection of plant leaf diseases using image
segmentation and soft computing techniques” Information Processing
REFERENCES In Agriculture 4 (2017) 41–49
[1] Saradhambal .G , Dhivya. R, Latha. S, R. Rajesh “Plant Disease [16] Alvaro Fuentes, Sook Yoon, Sang Cheol Kim and Dong Sun
Detection and Its Solution using Image Classification” International Park “A Robust Deep-Learning-Based Detector for Real-Time
Journal of Pure and Applied Mathematics. Volume 119 No. 14 Tomato Plant Diseases and Pests Recognition” Sensors 2017.
2018, 879-884 [17] SrdjanSladojevic, Marko Arsenovic, AndrasAnderla,
ISSN: 1314-3395 (on-line version) url: http://www.ijpam.eu DubravkoCulibrk and Darko Stefanovic “Deep Neural Networks
[2] Pallavi.S. Marathe“ Plant Disease Detection using Digital Image Based Recognition of Plant Diseases by Leaf Image Classification”
Processing and GSM” International Journal of Engineering Science Hindawi Publishing Corporation Computational Intelligence and
and Computing, April 2017, Website:http://ijesc.org/ Neuroscience, Volume 2016, Article ID 3289801, 11 pages
[3] RakeshChaware, RohitKarpe, PrithviPakhale, Prof.SmitaDesai“ http://dx.doi.org/10.1155/2016/3289801
Detection and Recognition of Leaf Disease Using Image Processing” [18] Sharada P. Mohanty, David P. Hughes And Marcel Salathe ” Using
International Journal of Engineering Science and Computing, May Deep Learning For Image-Based Plant Disease Detection” Front.
2017 , Website:http://ijesc.org/ Plant Sci., 22 September 2016 .
[4] Vishal Mani Tiwari&Tarun Gupta “Plant Leaf Disease Analysis [19] Sushil R. Kamlapurkar ”Detection of Plant Leaf Disease using Image
using Image Processing Technique with Modified SVM-CS Processing Approach” International Journal of Scientific and
Classifier” ResearchGate2017 Research Publications, Volume 6, Issue 2, February 2016 73 ISSN
[5] Xihai Zhang , (Member, Ieee), Yue Qiao , Fanfeng Meng, Chengguo 2250-3153
Fan , And Mingming Zhang “Identification of Maize Leaf Diseases [20] SrnderKumar, Rupinder Kaur “ Plant Disease Detection using
Using Improved Deep Convolutional Neural Networks” in Image Processing- A Review” International Journal of Computer
proceedings of IEEE June 26, 2018. Applications (0975 – 8887) Volume 124 – No.16, August 2015
[6] Guiling Sun , Xinglong Jia , and TianyuGeng “Plant Diseases [21] Ms. Kiran R. Gavhale, Prof. Ujwalla Gawande “An Overview of the
Recognition Based on Research on Plant Leaves Disease detection using Image Processing
Image Processing Technology” in proceedings of Hindawi, Journal Techniques” IOSR Journal of Computer Engineering (IOSR-
of Electrical and Computer Engineering, Volume 2018, Article JCE)7Volume 16, Issue 1, Ver. V (Jan. 2014)
ID 6070129, 7 pages.https://doi.org/10.1155/2018/6070129 [22] Pranita P. Gulve, Sharayu S. Tambe, Madhu A. Pandey, Mrs.
[7] M.N. Abu Bakar, A.H. Abdullah, N. Abdul Rahim, H. Yazid, S.N. S.S.Kanse.” Leaf Disease Detection of Cotton Plant Using Image
Misman and M.J. Masnan” Rice Leaf Blast Disease Detection Processing Techniques” IOSR Journal of Electronics and
Using Multi-Level Color Image Thresholding” in proceedings of Communication Engineering (IOSR-JECE) e-ISSN: 2278-2834,p-
Research Gate, Article· August ISSN: 2278-8735. PP 50-54 www.iosrjournals.org Innovation
2018.https://www.researchgate.net/publication/326800397 [23] Arti N. Rathod, Bhavesh A. Tanawala, Vatsal H. Shah “ Leaf
[8] Malti K. Singh, SubratChetia “Detection and Classification of Plant Disease Detection Using Image Processing and Neural
Leaf Diseases in Image Processing using MATLAB” International Network”International Journal of Advance Engineer ing and
Journal of Life Sciences Research ISSN 2348-3148Vol. 5, Month: Research Development (IJAERD) Volume 1,Issue 6,June 2014
October - December 2017, Available at: www.researchpublish.com [24] Guan Wang, Yu Sun, and Jianxin Wang “Automatic Image-Based
[9] Sujatha R, Y Sravan Kumar and Garine Uma Akhil “Leaf Disease Plant Disease Severity Estimation Using Deep Learning” Hindawi
Detection using Image Processing” Articlein Journal of Chemical Computational Intelligence and Neuroscience Volume 2017, Article
and Pharmaceutical Sciences · March 2017 ID 2917536, 8 pageshttps://doi.org/10.1155/2017/2917536
https://www.researchgate.net/publication/318109025
[10] Sujeet Varshney, TarunDalal “Plant Disease Prediction using Image
Processing Techniques- A Review”International Journal of Computer
Science and Mobile Computing, Vol.5 Issue.5, May- 2016, pg. 394-
398

0753