SMART AGRICULTURE MONITORING AND FERTILIZER

INJECTION USING AOT WITH DATA ANALYTICS

ABSTRACT
A proper agricultural support system is must for today's instantaneous world. The plants
are affected mainly due to diseases that are originated about 250 million years ago and the
improper supplement of nutrients to the soil at the appropriate time. To uncover the disease a
robust and practical technology i.e., Image processing is used. The system helps to find the
disease and sends the required fertilizers to the user through a mobile app. The Support Vector
Machine (SVM) based regression system helps to identify and classify the disease in plants. The
App displays basic parameters obtained from the sensors such as temperature and humidity. It
also acts as a bridge between user and the agricultural support system that paves way for making
the decisions quicker. The farmers can decide the mode of the fertilization injector according to
their need. The disease detection system and the soil monitoring system are interfaced with the
microcontroller.

INTRODUCTION
Agriculture
Agriculture is an important factor because it satisfies the basic needs of human. The term
agriculture is a science of cultivation, growth and raise of livestock. Over centuries, it also paved
a way for rise of civilization. In early days, people spent most of their lives in hunting wild
animals and gathering wild plants for the purpose of consumption. About 11,000 years ago,
people began to cultivate the cereals and root crops in their respective lands. In addition to the
farming, they took the shelter in the farming lands. The people learned to produce surplus foods.
In current scenario, the agricultural crops suffer from numerous reasons. Even though, there are
many advancement in technologies people failed to safeguard the potential value of agriculture
crops. The healthy nation lies in hands of healthy plants.
Need for Agriculture
Agriculture is a factor of survival for many people. About 70% of people directly depend on
agriculture for living. It is the primary source of income for many developing countries like India.
Agricultural sectors also provide fodder for domestic animals. The products like sugar, tea,
spices, tobacco, coffee etc constitute the major items of export. The export trade of a nation rely
largely on agricultural sectors. For example, agricultural products such as jute, tobacco, spices,
oilseeds, raw cotton, tea and coffee accounts for 18% of total value of export of a nation.
Problems in Agriculture
The agricultural sector has undergone numerous problems. The main problems include
improper nutrients, irrigation and poor growth. India is named as second largest irrigated country
in the world. Although it has received the name, the irrigation process is not assured due to
uncertain rainfalls. The salinity, water logging and alkalinity make the irrigation process much
tougher. The people are using Indian land for yielding over many years. This has led to depletion
and exhaustion of soil thus making lower productivity.
Manures and fertilizers can increase the agricultural production about 70%. The
utilization of rural and urban compost will provide manure to soil and remove waste from soil .

Role of plants
The humans depend upon plants either directly or indirectly. Over the years, humans used
seven thousand plant species as food. One of the important by-products of plant’s food making
process is oxygen. In order to survive, every living organism should intake oxygen. Plants also
help to provide shade for many animals, moderate the temperature and protect animals from the
excessive winds. It changes the rainfall pattern over large areas of earth’s surface. Plants add
nutrients to soil upon decomposing. This element beautifies the human world with its endless
green nature.

TECHNICAL DETAILS
Objective
In olden days, farmers found difficult to detect the disease accurately. Since, the
symptoms for most of the diseases are alike the chances of misinterpretation is also increased.
This led to injection of wrong fertilizer and damage to plants. In order to overcome this issue and
increase the production of crops, an agricultural system is developed.
BLOCK DIAGRAM

Figure 1 Block Diagram of Overall Proposed Solution

 The plants are subjected to disease due to many miscellaneous and environmental factors.
Once the disease is identified by image processing, the information regarding the disease and the
appropriate fertilizer is sent to the user through app. The farmer can select the fertilization mode
from the app.

Block Diagram Description

The images of the plant are captured and processed using image processing technique to
detect the disease of the plants. The sensor unit senses the parameters of soil. It detects the
temperature and moisture content of the soil. All these information are sent to the microcontroller
through HC-05 module and is displayed in LCD as well as in app. If the disease is detected the
disease information is send to the farmer along with the fertilizer suggestion from the Matrix
Laboratory (MATLAB) processing system to the controller using USB to TTL (Universal Serial
Port to Transistor-Transistor Logic) converter. Three modes are used to apply fertilizer to the
field. The farmer selects any one of the operational modes. The microcontroller will activate the
relay according to the mode given by farmer. By the data from the sensing unit, if the moisture
content of soil is low the irrigation is also provided to the soil. The Android App helps in making
the decision quicker and saves the time for the farmer with its operation from far places.
HARDWARE DESCRIPTION

Thermistor
The thermistor as shown in Figure 2 is a temperature sending element composed of
sintered semiconductor material that exhibits a large change in resistance in proportion to a small
change in temperature. Negative temperature coefficient thermistor exhibits a decrease in
electrical resistance when subjected to an increase in body temperature. It is capable of operating
over the temperature range of -100 ℃ to +600 ℃. Because of their very predictable
characteristics and their excellent long term stability, thermistors are generally accepted to be the
most advantageous sense for many applications including temperature measurement and control.

Figure 2 Diagram of Thermistor

Humidity Sensor
A humidity sensor as shown in Figure 3 senses, measures and reports both moisture and
air temperature. The ratio of moisture and air to the highest amount of moisture at a particular air
temperature is called relative humidity. It becomes an important factor, when looking for
comfort. It works by detecting changes that alter current or temperature in air. The amount of
water vapour needed to achieve saturation increases as the temperature increases. As the
temperature of a parcel of air decrease it will eventually reach the saturation point without adding
or losing water mass. The amount of water vapour contained within a parcel of air can vary
significantly.

Figure 3 Diagram of Humidity Sensor

Android of Things (AOT)

HC-05 module as shown in Figure 4 is an easy to use Bluetooth Serial Peripheral Port
(SPP) module, designed for transparent wireless serial connection setup. Serial port Bluetooth
module is fully qualified Bluetooth V2.0+EDR (Enhanced Data Rate) 3Mbps Modulation with
complete 2.4GHz radio transceiver and baseband. It uses CSR Bluecore 04-External single chip
Bluetooth system with CMOS technology and AFH (Adaptive Frequency Hopping). It has the
footprint as small as 12.7mm x 27mm. It simplifies overall design/development cycle.

Figure 4 Diagram of HC-05 module

SOFTWARE DESCRIPTION
MATLAB
MATLAB is a high-level language and the software act as an interactive environment
for numerical computation, visualization, and programming. Using MATLAB, one can analyze,
develop and create models and applications. The language, tools and built-in math functions
helps to derive multiple outcome ways and get a faster solution using programming languages
such as C/C++ or Java. MATLAB is used for applications such as signal processing and
communication, image and video processing, control systems and computational finance. More
than one million of engineers and scientists use MATLAB for computing.

MATLAB is known as Matrix Laboratory and the software uses vectors and matrices to
represent the data. Hence, the software is useful for solving linear algebra, algebraic and
differential equations and numerical integration. It also has powerful graphic tools to produce
nice pictures in both 2D and 3D.

Image processing
Image processing is a method to convert an image into digital form in order to extract
useful information from it. The input is image, like video frame or photograph and output may
be image or characteristics associated with that image. Usually, image processing system treats
images as two-dimensional signals. It is one of the rapidly growing technologies, with its
applications in various aspects of a business. It forms as a core research area within engineering
and computer science disciplines too. Image processing includes the following three steps: i)
Importing the image with scanner or photography, ii) Analyzing and manipulating the image
and iii) Output is either an image or report.
Purpose of Image Processing
The purpose of image processing areas follows visualization to observe the objects that
are not visible, image sharpening and restoration to create a better image, image retrieval to
seek for the image of interest, measurement of pattern to measures various objects in an image
and image recognition to distinguish the objects in an image.
Types of Image Processing
Analog
Analog or visual techniques of image processing are used for the hard copies like
printouts and photographs. Image analysts use various fundamentals of interpretation. The
image processing is not confined to area that has to be studied but on knowledge of analyst.
Therefore, the analysts apply a combination of personal knowledge and collateral data to image
processing.
Digital
Digital Processing techniques help to manipulate the digital images by using computers.
Since the data are collected from satellite platform, it contains deficiencies. In order to overcome
such flaws and to get original information, it has to undergo various phases of processing. All
data should undergo four different phases and they are preprocessing, enhancement and display,
information extraction.
Block Diagram of Image Processing
Algorithm in Image Processing
Image acquisition
Image acquisition in image processing is defined as the action of retrieving an image
from some source, usually a hardware-based source, so it can be passed through whatever
processes need to occur afterward. Image acquisition is always the first step in the workflow
sequence because without an image no processing is possible. The acquired image is
completely unprocessed and is the result of whatever hardware was used to generate it, which
can be very important in some fields to have a consistent baseline from which to work. The
overall block diagram of the image processing process is shown in Figure 5.

Input Images Image Pre-processing Feature extraction

Segmentation

Classification

Results

Figure 5 Block Diagram of Image Processing

Pre-processing
Pre-processing images commonly involves removing of low-
frequency background noise, normalizing the intensity of the individual particles, removing
reflections, and masking portions of images. Image pre-processing helps to enhance the data
before undergoing the processing.
Image segmentation
Image segmentation is nothing but a process of segmenting an image into multiple
partitions (sets of pixels, also known as super-pixels). The main aim of segmentation is to
simplify or change the representation of an image into something that is more meaningful and
easier to analyze. Image segmentation helpsto locate objects and boundaries (lines, curves, etc.)
in images. It is also seen as a process of assigning a pixel to each image. The pixels share the
same characteristics.
Feature extraction
Feature extraction nurses initial set of measured information and builds values
(features) which are supposed to be informative and non-redundant to facilitate the next
learning for better human interpretations. Feature extraction is related to dimensionality
reduction. When the input data is too large to be processed, it is suspected to be redundant (e.g.
the same measuring in each feet and meters, or the repetitiveness of images presented as
pixels). Then, it may be remodeled into a reduced set of options (also named a feature vector).
Determining a set of the initial options is termed Feature choice.
Image classification
Image classification helps to analyze the numerical properties of various image
features and organize data into categories. Classification algorithms employ two phases of
processing: training and testing. In the initial training phase, ccharacteristics of image features
are isolated. Based on this information a unique description of each classification category i.e.,
training class is created. In the subsequent testing phase, these features are used to classify
image features.
ANDROID APP
In this proposed system, Bluetooth electronics app is used for establishing serial connection
between HC-05 module and the microcontroller. It consists of various controls such as buttons,
switches, sliders, pads, lights, gauges, terminals, accelerometers and graphs. This App is used to
control devices. The communication is done through Bluetooth HC-05 module. It helps to
select the mode of the system. The buttons are used to change the mode. Indicators are used to
display the mode. Gauges are used to measure and display the temperature and humidity of the
field. In the pads status of the device is displayed.
In this project, thermistor (10k) is used to measure the temperature of the soil. The
thermistor acts as a transducer by converting the measured heat energy into electrical energy.
Since this thermistor belongs to the type of negative temperature coefficient, the resistance
decreases with increase in temperature. The range of thermistor ranges from 55℃ to +125℃.
One end of the thermistor is connected to the 9th pin (PORT E) of PIC microcontroller and other
end to the supply as shown in Figure 4.
Interfacing Diagram of Thermistor with PIC16F877A

Figure 6 Interfacing Diagram of Thermistor with PIC16F877A

Interfacing Diagram of LCD with PIC16F877A

Figure 7 Interfacing Diagram of LCD with PIC16F877A

Here, 16x2 LCD is used to display the parameters of soil which are measured using
thermistor and humidity sensor. About 32 characters are displayed in two lines with each 16
characters. The basic supply to turn on the LCD is 5V. From the below Figure 7, the data bus
from 11- 14 are used for transferring 4 to 8 bit data and they are attached to pins from 37-40 of
the controller. LCD pins from 7-10 are not used because they can transfer only 4 bit data. In order
to adjust the colour contrast variable resistor 103 is connected to Vee pin of the LCD. Vss is
connected to 31th pin and Vdd is connected to 32th pin of controller.

Interfacing Diagram of Pump Motor with PIC16F877A

A pump as shown in Figure 8 is a mechanical device which helps to move the fluids. In order
to perform some work pump needs some energy. Hence, a motor or engine is used for converting
the electrical energy into mechanical energy. Both pump and motor plays an important role in the
process. Here, a motor which supports 165V-250V with 50Hz is used for pumping action. A relay is
an electromagnetic switch which operates at lower current. It helps to drive bigger apparatus with
small amount of current. In this project, microcontroller will output a constant voltage of 3.3V.
Since, it is very low to start a motor relay is used. The relay operates with 50-60 Hz at DC 6V.
When the current passes through the relay an electromagnetic field is developed within the coils and
it attracts the common terminal to and makes the circuit to get closed. Now, the parallel motor
circuit will also get turned on. The supply to the motor is given through the step down transformer.
Two relays are used one for fertilizer and other for water. The common terminal of both relays and
the negative terminal of the transformer are connected to supply. The non closed terminal of both
relays is connected with one end of pump and other end of pump and positive terminal of the
transformer to the supply.

Figure 8 Interfacing Diagram of Pump motor PIC16F877A

Interfacing Diagram of Bluetooth Module with PIC16F877A
HC-05 module mentioned in Figure 9 is used for transferring information from
microcontroller to the android app. The frequency of Bluetooth is 2.4 GHz. The range of
Bluetooth is less than 100m. The Vcc pin of the Bluetooth is connected to 5V supply. The
Bluetooth operates in command mode. The module communicates with the help of USART at
9600 baud rate with the microcontroller. The 25th transmission pin of controller is connected to
receiver pin of module. Also, the transmission pin of module is connected to 26 th receiver pin of
controller. The state pin helps to check whether the Bluetooth works properly or not. The
Bluetooth works under serial communication protocol. The information regarding the
parameters of soil such as temperature and humidity are passed to the mobile device through
Bluetooth. Also, the fertilizer ratio and the process of fertilization is controlled from module
and sent to controller.

Figure 9 Interfacing Diagram of Bluetooth Module with PIC16F877A

Image Processing of Leaf

Image processing converts an image into digital form and performs required operations in it, in
order to get an enhanced image or to extract useful information. The main steps in detecting
disease are Image acquisition, image pre-processing, feature extraction, segmentation and
classification. The leaves are one of the main parts in plants. The plants can be classified into
different types by analysing the morphological and genetic characteristics of the leaf. In this
project, leaves are taken to identify the disease. Image acquisition is a process of retrieving an
image from the source, which is probably a hardware-based source. Here images are taken from
the database and directly from the field. The retrieved images are resized into 300 x 300-pixel
size. Next, images are subjected to image pre-processing phase to suppress the noise and
enhance its quality. The pre-processing method integrates the combination of image
enhancement, colour conversion, image resizing and filtering of an image. The RGB images are
converted into the grey images using colour conversion. The canny operator is used to eliminate
the noise without affecting the features of the image. Hence, the edges of the object will be
clear and precise. The features such as colour, size and shape are taken from the image to
segment it.

RESULTS AND DISCUSSION

The infectious and the healthy leaves are provided as the sample for the processing unit.
Initially the natural image of leaf is provided as the input to the system. The image is involved
in pre-processing, segmentation and the features are identified from the image. If the feature
value of the test image varies from the original image, the disease detection is performed. The
name of the disease along with the fertilization is displayed to the user.

Figure 10 Image Processing Page

The natural image to be tested is loaded in the resultant place of the page and the
reference image is taken from the image database as shown in Figure 11.

Figure 11 Identification page a) Plant Dryness

 Figure 12 b) Plant Normal

Figure 13 c) Disease Occurrence

The plant is checked for the following three conditions as shown in Figure 11 a) plant
dryness, Figure 12 b) plant normal and Figure 13 c) disease occurrence in the image processing
unit and it is displayed on the app.
Auto Mode
Once this mode is selected, the respective fertilizer ratio for the plant is automatically
taken from the program and the relay triggers the fertilizer pump. The information regarding the
temperature and the water content is displayed in the app as shown in Figure 14 as well as in
LCD.

Figure 14 Snapshot of Auto Mode

Manual Mode
Once this mode is selected, the farmer can input the fertilizer ratio through the mobile
app as shown in Figure 15. This mode simply activates the fertilizer pump through the relay.
The farmers can fill the tank with the fertilizer of their choice.

Figure 15 Snapshot of Manual Mode

Smart Mode
Once this mode is selected, the MATLAB unit is processed and the respective disease
along with the fertilizer is sent to the user app through the Bluetooth module. The parameters
such as temperature and humidity are shown in Figure 16.

Figure 16 Snapshot of Smart Mode

 Figure 17 Overall Diagram of Proposed Agricultural System

The overall agricultural system is mentioned in the Figure 5.6. The plastic container acts as
a tank. One of the tanks is used for injecting fertilizer and other is used for injecting water. The
system is controlled through the microcontroller.

CONCLUSION
India’s economy still mainly depends upon agriculture. Most of our population lives in
the rural areas, so farmers are not exposed to technology and automation to a greater extent.
The leaves are segmented using canny edge detection. After the completion of segmentation
process, various trained and test images are compared using SVM classifier. The connection
between microcontroller and MATLAB is established using USB to TTL converter. From the
microcontroller the data are displayed on the app through Bluetooth module. The parameters
from soil such as moisture and temperature are measured and displayed on the LCD as well as
in App. The fertilization process works in three different modes according to the convenience
of the user. The disease information is displayed on the App. Since the quality of the plant is
increased automatically the productivity of the crop is also enhanced. Thus, the agricultural
sector will earn more profit.
REFERENCES
[1] M. K. Gayatri, J. Jayasakthi and Dr. G. S. Anandha Mala, “Providing Smart
Agriculture Solutions to Farmers for better yielding using IOT,” IEEE
International Conference on Technological Innovations in ICT for Agriculture
and Rural Development, 2015.
[2] G. Rajesh Kumar, “Crop Selection Method to Maximize Crop Yield Rate using
Machine Learning Technique”, International Conference on Smart Technologies and
Management for Computing Communication Controls Energy and Materials, 2015.
[3] S. R. Kamlapurkar, “Detection of Plant Leaf Disease Using Image
Processing Approach,” International Journal of Scientific and Research
Publications, vol.6, no.2, pp.73 to 76, 2016.
[4] K. Thangadurai and K. Padmavathi, “Computer Image Enhancement for Plants
and Leaves Disease Detection”, World Congress on Computing and Communication
Technologies, 2014.
[5] Dashsonali K, Chiranjeevi U and R. Jena Trinadh akula, "Comparative Study of
Image Texture Classification Technique", International Conference on Electrical
Electronics Signals Communication and Optimization IEEE, 2015.
[6] F.Susan and E.S.E.A.K. Ellakwa, “Integrated Ontology for Agriculture Domain”,
International Journal of Computer Applications, vol.54 no.2, pp.46 to 53, September
2012.
[7] J. Gutirrz, F. Villa-Medina, A. Nieto_Garibay and M. Porta_Gndara,
“Automated Irrigation System using a Wireless Sensor Network and GPRS
Module”, IEEE Transactions on Instrumentation and Measurement, vol.63, pp.166
to 176, Jan 2014.
[8] Saurabh Malgaonakr, Sanchi Soral, Shailja Sumeet and Tanay Parekhji, “IEEE
Study on Big Data Analytics Research and Development” Mulund, Mumbai, India,
vol.65, pp.187 to 196, March 2015.
[9] Baljitkaur and Dilipkumar, “Development of Automated Nutrients Composition
Control Fertigation System, “International Journal of Computer Science,
Engineering and Application, vol.3, June 2013.
[10] Isabelle M.Carbonell, “Ethics of Big data in Agriculture”, Internet Policy Review,
2016.