IRJET- Sugarcane Disease Detection and Controlling using Image Processing

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 06 Issue: 04 | Apr 2019 www.irjet.net p-ISSN: 2395-0072
© 2019, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1586
Sugarcane Disease Detection and controlling Using Image Processing
Kranti Deshpande1, Sayali Gadekar2, Gayatri Patil3, Prof. Vijay Bhong4
1,2,3Student, Dept. of ENTC Engineering, SVERI’s College of Engineering, Pandharpur, Solapur University, Solapur,
Maharashtra, India
4Assistant Professor, Dept. of ENTC Engineering, SVERI’s College of Engineering, Pandharpur, Solapur University,
Solapur, Maharashtra, India
----------------------------------------------------------------------***---------------------------------------------------------------------
Abstract:- India is the fast growing country and agriculture
is the main source for the countries development. Due to
industrialization and globalization concept the fieldisfacing
hurdles. Now a day’s technology plays a vital role in
agriculture. Identifying the health of sugarcane plays an
important role for successful cultivation. Identifying plant
disease wrongly lead to huge loss ofyield,moneyandquality
of product. Generally we can observe symptoms of leafs,
flowers, stems and fruits etc. So here we use leafs for
detection of disease.
Keywords:- Plant disease, Image acquisition, pre-processing,
Feature extraction, symptoms, and pesticides.
1. INTRODUCTION
Agriculture is important field as it provides food.
The old and classical approach for detection andrecognition
of plant diseases is based on necked eye observation. In
some countries, consulting experts to find out plant disease
is expensive and time consuming due to availability of
experts. India experiences /variety of climates ranging from
tropical in the south to the temperature in the south. Due to
unpredictable climate changes .Thereislack ofnutrients and
minerals to the crops. This lead to deficiency diseases which
in turn affect the crop productivity. Plants get affected by
micro- organisms like fungi, virus and bacteria.
Recognition of plant disease can be effectively done
through leaves as they are the conspicuous and delicate
piece of a plant. Automatic detection of plant disease is
fundamental to identify the symptoms of diseases in
beginning times when they show up on the developing leaf
and product of plant.
The object of this paper is to concentrate on the
plant leaf detection based on texture color of the leaf. Leaf
presents several advantages over flower. There are four
section 1] Introduction of sugarcane diseases, plant leaves
analysis, various type of leafdiseases.2]Discussiononrecent
work carried out in this area.3] Basic methodology for leaf
disease detection.4] Conclude topic along with possible
future directions.
1.1 Plant disease analysis and symptoms:
Imag analysis can be applied for the following purpose:
1. To detect plant disease.
2. To detected affected are by disease.
3. To find boundaries of the affected area.
4. To find color of affected area.
There are some common symptoms fungal, bacterial and
viral.
1.1.1 Bacterial disease symptoms:
These disease caused by tiny pale green spots on leaves.
Under dry condition spots have a speckled appearance. The
shape, size and function of the leaf vary.
1.1.2 Fungal disease symptoms:
These disease characterized by lower olderleaveslikewater
soaked, gray green spots on leaves. When fungus get spread
this spots darken and then white fungal growth get form.
1.1.3 Viral disease symptoms:
Among all diseases, disease caused by virus is difficult to
identify. Leafhoppers, aphis common carries these diseases
e.g. Mosaic virus.
2. LITRATURE SURVEY
In paper [1] authors focused The application of texture for
detecting the sugarcane disease has been explained by color
transformation structure RGB is converted into HSV space
because HSV is a good color descriptor. Masking and
removing of green pixels with pre-defined threshold level
and segmentation is done.
In paper[2] considered captured leaf images are segmented
using k-mean clustering method to form clusters. Features
are extracted before applying classification technique.
In paper [3] authors used lab view and MATLAB for
detection of plant disease. Detectionin earlystageispossible
due to the MATLAB. Morphological operationsdonewiththe
help of image pre -processing.

In paper [4] Introduced RGB images havebeenconcentrated
into Hue Saturation and Hue (HSV) color space
representation and showed the H, S and V components.
In paper [5] Presented an investigation of leaf spot disease
using image segmentation c-means classification. Median
filter used for image enhancement.
In paper [6] Authors introduced techniquewithimprovedk-
means method. After acquisition RGB converted into grey
scale and HSV.In threshold process histogram and multi-
level threshold are obtained to isolate relevant image.
Centroid value is calculated.
In paper [7] Feature extraction color, morphology and
feature are extracted and filter is used in texture and
morphology for obtaining boundary of image.
3. WORKING METHODOLOGY
The block diagram of proposed methodology is as below:
Fig -1: Flow chart
In this proposed methodology we made a MATLAB stand -
alone application .
A. Image acquisition:
It is the first method of image processing
It is described as capturingimagethroughdigital camera and
stores it in digital media for further matlab operations.
Fig- 2: Original image
B. Pre-Processing:
The main purpose of image pre-processing is to
improve the image data containedunwanteddistortion. Pre-
processing utilizes different systems for example changing
image size and shape, filtering of noise, image conversion,
and morphological operations.
C. Image segmentation:
In this project we used segmentation techniquesoinput
image is divided into three clusters for good segmentation
result.
1. Region based
2. Edge based
3. Threshold based
D. Feature Extraction:
Feature extractions describe the relevant shape
information contained in pattern. Transforming input data
into the set of features is called feature extraction.
Processing involve following steps:
1. Image acquisition
2. Conversion of RGB into HSV image.
3. Segmentation of the components.
4. Obtaining the useful image from the segmented
image.
5. Color co-occurrence method.
K means clustering algorithm:
This algorithm is used to cluster/divide the object
based on the feature of the leaf in to K number of groups.
The algorithm of K means
1. User should select the value of K.
2. Every pixel is assigned to its nearest centroid (K).
3. The position of centroid is changed by means of data
values assigned to the groups.
Out of these three clusters classification is done for only
cluster which has affected area.
Color analysis:
Color image processing and analysis is concerned
with the manipulation of digital color images. HSI model is
proposed to improve the RGB model. The process of
selecting the best color representation involves knowing

how color signals are generated and what information is
needed from these signals.
1. Input Image
Fig-3: Original image
2. Output Images
Result-1(Range)
Fig -4: Output Image
We got this output image by calculating the value here the
Red root disease is found and use Copper Oxychloride.
Result-2(color)
Fig -5: Output Image
This is the output image by Color feature extraction on
image. The White color showing the presenceofdisease.Red
root disease found spray Oxychloride
Result-3(Range)
Fig-6: Output Image
This is the output image of pixel range. Here Red
root Image is found use copper Oxychloride.
4. RESULT
The output gives the information of the disease of
the sugarcane plant along with its pesticides. By performing
pixel range code the result is more efficient than the color
extraction feature.
5. CONCLUSION
This project is very useful for farmers and laboratory
where they can easily protect their sugarcane and there will
be increase in growth of production. The main focus of this
project is to help the farmers, distress from loss due to
imperfect information of a choice of disease.
6. FUTURE SCOPE
In future, the proposed system would be deployed and
tested in real field. User friendly app could be developed to
alert the farmers and provide them with solution in real
time.
REFERENCES
[1]. Sandesh Raut and Amit Fulsunge, “Plant Disease
Detection using Image Processing
MATLAB”International Journal ofInnovativeResearchin
Science, Engineering and Technology vol.06,June-2015.
[2]. Surendar Kumar and Rupendar Kaur, “Plant Disease
Detection Using ImageProcessing”International Journal
of Computer Application vol.124, August-2016.
[3]. Kiran R. Gavahale and Ujwala Gawande,” Research on
Plant leaf Disease Detection Using Image Processing”
International Journal of Computer Application vol.16,
Jan 2017.
[4]. Wabale A. A. and Mulajkar R. M., ”Plant Health Detection
using Image Processing Robot” vol.02, June-2017.

[5]. Thamil Azhagi, Swetha, ”Plant Pathology Detection and
Control”, International journal of Engineering sciences
Rsearch Technology, vol.03,March-2017.
[6]. Shweta R. Astonkar, V. K. Shyandilya, “Detection and
Analysis of Plant diseaseusingImageProcessing”vol.04,
Apr-2018.
[7]. Vignesh Dhandapani, S. Rema ,” Plant Health Monitoring
Using Image Processing”, International journal of
engineering research in Computer Science and
Engineering, vol-05, March-2018.
[8]. Rafael C. Gonzalez,” Digital Image Processing”.

IRJET- Sugarcane Disease Detection and Controlling using Image Processing

Recommended

More Related Content

What's hot (20)

Similar to IRJET- Sugarcane Disease Detection and Controlling using Image Processing (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Sugarcane Disease Detection and Controlling using Image Processing