Plant Monitoring using Image Processing, Raspberry PI & IOT

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 10 | Oct -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 1337
PLANT MONITORING USING IMAGE PROCESSING,
RASPBERRY PI & IOT
Prof. Bhavana Patil1,Mr. Hemant Panchal2, Mr. Shubham Yadav3 , Mr. Arvind Singh4,
Mr. Dinesh Patil5
1Professor, Dept. of electronics and telecommunication, A.C.Patil College, Navi Mumbai, Maharashtra, India
2,3,4,5 Student, Dept. of electronics and telecommunication, A.C.Patil College, Navi Mumbai, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In India, 70% of its people are in the
Agriculture sector. The remaining 30% of people are an
inseparable part of this field. Because from agriculture, we
get various raw materials and especially crops, which serve
as a staple food for people. But The crops and plantations
get destroyed mainly due to two major reasons, first reason
is destruction by natural calamities such as flood,
earthquake, drought, famine, etc. and second reason is the
destruction by the pathogens.98% of the destruction is
caused by the pathogens and 2% of the destruction happens
due to natural calamities. So the need for the plant disease
detection was felt. The traditional methods were inaccurate
and not effective. So various researches in this field lead to
inclusion of image processing for accurate detection of
disease by using plant leaf. Various spots, patterns on plant
leaf are useful in detecting the disease. Further
advancement was use of digital image processing for more
accurate results. On referring various reputed IEEE,
international conference and international journal papers
regarding this field, it was found that none of these suggest
remedy for curing the plant disease.
Key Words: component; formatting; style; styling;
1. INTRODUCTION
In India, Farmers have a great diversity of crops. Various
pathogens are present in the environment which severely
affects the crops and the soil in which the plant is planted,
thereby affecting the production of crops .Various disease
are observed on the plants and crops .The main
identification of the affected plant or crop are its leaves.
The various coloured spots and patterns on the leaf are
very useful in detecting the disease. The past scenario for
plant disease detection involved direct eye observation,
remembering the particular set of disease as per the
climate, season etc. These methods were indeed
inaccurate and very time consuming. The current methods
of plant disease detection involved various laboratory
tests, skilled people, well equipped laboratories etc. These
things are not available everywhere especially in remote
areas. The method which I am implementing insures the
automatic detection of crops. The very first
implementation of plant disease detection using image
processing was done by shen weizheng, wu yachun, Chen
zhanliang and Wei hongda in their paper [9].The method
which I am implementing makes use of digital image
processing for plant disease detection.
2. LITERATURE SURVEY
Let’s have a brief overview of the various papers, which I
have referred for implementation of my project. The usage
of image processing technology for plant disease degree
grading eliminates the subjectivity of traditional
classification methods and human-induced errors. Thus
the estimation credibility is improved and accurate data
are provided for disease studies. The method is also
convenient, which simply needs computers, digital
cameras with the Combination of necessary software
programs to realize for the disease batch grading. The
accurate detection and classification of the plant disease is
very important for the successful cultivation of the crop
and this can be done by using image processing. The basic
steps for disease detection using image processing include
image acquisition, image pre processing, feature
extraction, detection and classification of plant disease.
Enhanced images have high quality and clarity than the
original image. Colour image have primary colors red,
green and blue. It is difficult to implement the application
using RGB because of its range. Hence they convert RGB to
grey images. Detection of plant disease through some
automatic technique is beneficial as it reduces a large
work of monitoring in big farms of crops and at very early
stage itself detects the symptoms of the disease. They have
presented a survey on various classification techniques.
Abdul hallis et al in their paper, have used MATLAB for
feature extraction and image recognition. Here digital
camera is used for image capturing. Mrunalini and
Prashant. R.Deshmukh Compares Otsu threshold and the K
means clustering algorithm for infected leaf analysis. The
clarity of k means clustering is more accurate than other
method. J.K.Patil in his paper describes the possible
approach for extraction of low level image features like
color and texture. Anand kulkarni in his paper discusses
the Gabor filter and ANN for feature extraction and
classification respectively . Haiguang Wang in his paper
stresses on the principal component analysis. PCA could
reduce the dimensions of the obtained data under the
premise of retaining the total data information, reduce the

nod .of neurons in the input layer and increase the speed
of neural networks. Maintaining the Integrity of the
Specifications
3. STEPS FOR PLANT DISEASE DETECTION
3.1 IMAGE ACQUISITION
The images of the plant leaf can be acquired using two
ways. First way is to capture image using the external
camera, here we have used iball web camera and second
way is to get the image from the email etc. The input image
is converted to color space.
Fig.(3-1) image acquisation
3.2 IMAGE PRE-PROCESSING
To remove the noise in the image, different pre processing
techniques are considered such as image clipping for
obtaining the concerned area by cropping; image
smoothing is done to reduce in the obtained image by
using the smoothing filter. Image enhancement is done to
increase the contrast.
Fig.(3-2) image cropping example
Fig.(3-3) image smoothing
Fig.(3-4) enhanced image of leaf
3.3 IMAGE SEGMENTATION
Segmentation means partitioning of images into various
part of the same feature or having some similarity. The
segmentation can be done using Otsu’s method, K means
clustering, converting RGB to HSI model etc. The K means
clustering is used for classification of the object based on a
set of the features into K no. of classes. The classification of
the object is done by minimizing the sum of the square of
the distance between the object and the corresponding
cluster.
Fig.(3-5) image segmentation

3.4 FEATURE EXTRACTION
Feature extraction plays an important role for
identification of an object. In many application of image
processing, feature extraction is used. T he features which
can be used in plant disease detection. In paper [3],
Monica jhuria et al considers color, texture and
Morphology as a feature for disease detection. They have
found that morphological result gives better result than
the other features. Texture means how the color is
distributed in the image, the roughness, hardness of the
image. It can also be used for the detection of infected
plant areas. Color, texture, morphology, etc. are the
features which can be used in plant disease detection. In
my project for feature extraction, we have converted the
image to grey scale. The image which is in the color space
is converted into grey image using,
F(x) =0.2989*R+0.5876*G+0.114.*B.
Fig. (3-6) Feature extraction
3.5 CLASSIFICATIONS
For the classification and the detection, we have used SVM
(support vector machine) with .mat file. The SVM is the
supervised learning model with associated learning
algorithm that analyzes data used for classification and
regression analysis. A SVM constructs a hyper plane or a
set of hyper planes in a high or infinite dimensional space
which can be used for classification, regression or other
tasks. Intuitively, a good separation is achieved by the
hyper plane that has the largest distance to the nearest
training-data point of any class (so-called functional
margin), since in general the larger the margin the lower
the generalization error of the classifier. The.mat file store
the various features of the affected leaf.
4. BASIC TYPES OF PLANT FAMILIES
4.1 MONOCOT FAMILY PLANT
These type plants have one seed leaf. The vein structures
of Monocot family plants are parallel and straight wheat,
ginger, corn, rice, millet.
4.2 DICOT FAMILY PLANT
These type plants have two seed leaves. The vein
structures of Dicot family plants are nested and complex
structure. Cotton, potatoes, tomatoes, beans, honeysuckle,
roses.
5. FRUIT SIZE DETECTING AND GRADING
5.1 PROCESSING FLOW
Fig (5-1) fruit image size detection and grading
Take apple as the processing example, according to, the
apple size is its diameter, which is the longest distance in
the apple’s cross section. So the detecting program is
focused on how to calculate the diameter in an apple side
view image. The fruit image size detecting and grading
processing flow.
5.2 COLOR DETECTION
In the process of fruit color is detected according toRGB
values; here fruits are sorted according to color and size.
So for e.g. two fruits are considered say tomato having red
color and guava having green color, so in this step work is
going to find out color of a fruit by using RGB values of an
image taken from the camera, this image can be processed
by using MATLAB software and accordingly color can be
detected i.e. green or red.
Color detection algorithm:
1) Start
2) Read the input color image using imread function.
3) Read the input pixel of color image in three different
planes (RGB) and store it into three variable r, g, and b.
4) Read the small region of fruit to detect color of fruit.
5) Store in different variable r1, g1, b1.
6) Calculate the mean of r1, g1, b1 and store into variable
r2, g2, b2.

7) Compare the value with threshold.
8) If g2>threshold, Color detected is green.
9) If r2>threshold, Color detected is Red.
10) End.
5.3 EDGE DETECTION
Once color is detected, there is a need to find out size of a
fruit. The size of circular shaped fruit is its diameter. The
edge extraction is key factor for size detecting. After gray
image, the most powerful edge-detection method that
finds edge is the canny method.
The Canny method differs from the other edge-detection
methods in that it uses two different thresholds (to Detect
strong and weak edges) and includes the weak edges in
the output only if they are connected to strong edges. This
method is therefore less likely than the others to be fooled
by noise, and more likely to detect true weak edges.
5.4 FRUIT SIZE DETECTING ALGORITHM
In order to calculate this diameter, the fruit’s natural
symmetry is considered, so the fruit size detecting
algorithm based on its symmetry mainly contains two
parts: finding the center coordinates of fruit’s shape in
image and finding fruit’s axis in image . The algorithm is
described as follows:
Finding the center coordinate of fruit’s shape in image:
The center coordinate can be easily calculated once
finding the edge sequence points. Suppose the finding
edge sequence point is q (xi, yi), i=1, n. the center
coordinates of fruit’s shape is (cx, cy).
6. PLANT MONITORING SYSTEM OVERVIEW
Fig (6-1) monitoring system
Shown in figure , components of the automatic plant
monitoring system on Raspberry Pi Model-B and Arduino
Uno are demonstrated. The system can notify using a real-
time alarming system to smart phones reporting such as
the current and daily highest/lowest temperature,
humidity, and weather quality of the farm surroundings.
Users can also control the filter fan switches and
customize the notification system to the smart phone.
7. MONITOURIN SYSTEM METHODOLOGY
7.1 RASPBERRY PI
Raspberry Pi is a small computer board working on the
Linux operating system which connects to a computer
monitor, keyboard, and mouse. Raspberry Pi can be
applied to a electronic structure and programming
network work, it can also serve as a personal computer
and Apache Webserver, MySQL could be installed in the
board.
A GPIO pin can be used as either a digital input or a digital
output, and both operate at 3.3V. Unlike the Arduino, the
Raspberry Pi which does not have any analog inputs. For
that you must use an external analog-to-digital
converter(ADC) or connect the Pi to an interface board
must be used.
7.2 ARDUINO
Arduino is an open-source microcontroller compatible
with developed platforms. The controller appears not to
be expensive and uses low electrical power, 5.5 volts. C
and C++ were employed for this development. Arduino can
connect to a computer via the Universal Serial Bus (USB)
and perform with compatible connected accessories in
both analog signal and digital signal. The Arduino is a
microcontroller platform, mounted on a board that plugs
easily into most computers. It allows the user to program
the onboard Atmega chip to do various things with
programming language, in programs called sketches.
7.3 HUMIDITY SENSOR MODULE
Environmental conditions directly affect animal livelihood
contributing to some chronic epidemics such as Bird Flu
and Hand Foot and Mouth Disease. Therefore, DHT22 is
use as a censor for measuring temperature (for both
Fahrenheit and Celsius value) and humidity. The
measurement unit will be demonstrated in a digital signal
form.

7.4 GAS SENSOR MODULE
The module works as a Air Quality Detection Gas Sensor,
this is sensitive to gas dangerous to human, applied to
measure NH3, NOx, Alchohol, Benzene, CO, and CO2. The
module is also used for controlling weather conditions and
air cleaners in buildings. The measurement unit is
presented in a analog signal.
In this research, three censor forms were used, consisting
of MQ-2, MQ-135, and MQ-136 because of its gas
measurement differences.
7.5 PHOTOSENSITIVE SENSOR MODULE (LDR)
A light sensor was used for measurement of light intensity
especially for naked eye light, its unit is called Lux . Light
Dependent Resistor (LDR) is a light sensitive resistance
changing electronic resistance when there is a light
incidence, called Photo Resistor or Photo Conductor. The
resistor was made from Semiconductor, Cadmium Sulfide
(Cds) or Cadmium Selenide (CdSe). These two substances
are semiconductors coated in a ceramic sheet as a base.
7.6 HARDWARE CONNECTION
The Raspberry Pi and Arduino were connected via UART.
The connection was a serial communication as Full Duplex
since there was two-ways that data could be transmitted
via pin TX and RX. A direct connection between the
Raspberry Pi and Arduino was prohibited, because of its
electrical potential differences, which is 3.3 volts for the
Raspberry Pi and 5 volts for the Arduino. Bi-directional
Logic Level Converter should be used to separate them. A
connection between a camera and Raspberry Pi by
applying Common System Interface (CSI) serves as a point
to point connection, providing a fast data transmission and
low energy consumption. MJPG-Streamer is a basic
program command copying data from a single input to
multiple outputs. A photo could be presented in a network
system accessing from a web browser on a computer. In
this study, a photo from a camera would be taken to
demonstrate on a smart phone. All sensors would be
connected via Board Arduino and the data would be
transmitted from UART to Board Raspberry Pi. Raspberry
Pi works as a controller of a ventilator, notifying a working
condition to the smart phone and served as a data sender
to store in a server computer.
7.7 SOFTWARE
Fig (7-1) Flowchart of programming in Raspberry Pi
A connection between Raspberry Pi and smart phones was
investigated in this study. Programming of a
communication between a server and a client consisted of
two perspectives. The Raspberry Pi would check whether
there was any connected clients, if so, a mutual data
transmission would be done. By doing this, an application
would send the data via Socket referring to IP Address and
Port in Transport Layer using TCP protocol.
Python programming is applied for the development of
Raspberry Pi. Python would read the Arduino signal value
via UART and then collect the obtained signal to the

database for processing. If the value surpassed threshold,
the over signal would be sent to GPIO pins to aware the
analog signal.
3. CONCLUSIONS
The accurate detection and classification of the plant
disease is very important for the successful cultivation of
the crops, this can be done using digital image processing.
In this project, the detection as wells the remedy for
curing it is achieved. This project utilizes GSM so as to
send the message to every kind of mobile handset. This
project utilizes various image processing techniques
which provide accurate results.
The proposed system is a demo version, so for a large
scale production the number of cameras and length of
conveyor system can be modified. This work presents new
integrated techniques for sorting and grading of different
fruits. Generally image capture is a big challenge as there
is a chance of high uncertainty due to the external lighting
conditions, so we are taking the advantage of gray scale
image which are less effected to the external environment
changes as well as beneficial for finding size of a fruit.
Same way while collecting fruit from conveyor system by a
main plate there is variation in the weight measurement of
a fruit so further design can be modified so fruits can be
collected stably. Speed and efficiency of a system can be
further improved by using ARM9 or ARM11 processor for
the same purpose.
REFERENCES
1. Sachin .D.Khirade, A.B.patil,” Plant disease
detection Using image processing,”2015,
International conference on computing
communication control and automation, IEEE. J.
Clerk Maxwell, A Treatise on Electricity and
Magnetism, 3rd ed., vol. 2. Oxford: Clarendon,
1892, pp.68-73.
2. Vijai singh, Varsha, A.K.Mishra”Detection of
unhealthy region of plant leaves using image
processing and genetic algorithm”, 205, ICACEA,
India. K. Elissa, “Title of paper if known,”
unpublished.
3. Monica Jhuria, Ashwani kumar and Rushikesh
Borse, ”Image processing for Smart farming,
detection of Disease and Fruit Grading,”
proceeding of the 2013, IEEE, second
international conference on image Information
processing.
4. Patil.J.K,Raj kumar,”Feature Extraction of diseased
leaf images 2012,journal of signal and image
processing.
5. Hongshe Dang, Jinguo Song, Qin Guo, “A Fruit Size
Detecting and Grading System Based on Image
Processing,” 2010 Second International
Conference on Intelligent Human-Machine
Systems and Cybernetics,pp83-86

Plant Monitoring using Image Processing, Raspberry PI & IOT

More Related Content

What's hot (20)

Similar to Plant Monitoring using Image Processing, Raspberry PI & IOT (20)

More from IRJET Journal (20)

Recently uploaded (20)

Plant Monitoring using Image Processing, Raspberry PI & IOT