Sunteți pe pagina 1din 3

2019 5th International Conference on Advanced Computing & Communication Systems (ICACCS 2019)

Smart Leaf Infection Identification and Fertilizer


Spray
M. Azlan Ahamed Habib S.Aakash
B.E Electrical and Electronics Engineering B.E Electrical and Electronics Engineering
Sri Krishna College of Technology Sri Krishna College of Technology
Coimbatore Coimbatore
15tuee028@skct.edu.in 15tuee001@skct.edu.in
M. Harisudhan T.Bharani Prakash
B.E Electrical and Electronics Engineering Assistant Professor
Sri Krishna College of Technology Sri Krishna College of Technology
Coimbatore Coimbatore
15tuee051@skct.edu.in bharaniprakash.t@skct.edu.in
S.Aaron
B.E Electrical and Electronics Engineering
Sri Krishna College of Technology
Coimbatore
15tuee002@skct.edu.in

Abstract: In this paper an automated system has been proposed system is to detect plant diseases using IoT.
developed to determine whether the plant is normal or Hence, in the proposed work we have considered detection
diseased. The normal growth of the plants, yield and of plant disease present on leaves. The discrimination of
quality of agricultural products is seriously affected by normal and affected plant leaf can be measured based on
plant disease. This paper attempts to develop an variation in temperature, humidity and colour. Image
automated system that detects the presence of disease in processing module (called as IP-module) with robotic setup
the plants. An automated disease detection system is will move over the land. First the camera is enabled then its
developed using sensors like temperature, humidity and start to capturing the plant leafs. Then these images are
colour based on variation in plant leaf health condition. processed in order to Pre-processing, Feature Extraction,
Here, IOT comes into picture to supply semi-automated
Segmentation and Classification. The ip-modules details
or fully-automated frameworks to supply superior
further forward through internet to end-users for analysis if
comes about.
any. After completing the identification process, a spray
Keywords: Internet of Things (IoT), Image fertilizer module will spray & alert the end-user if disease
Processing, Disease Diagnosis, Arduino, Robotic Setup identified otherwise it moves to another plants.

I. INTRODUCTION II. METHODOLOGY


Tropical countries like India, Agriculture is main In this project we are design the automatic robot module
productivity & development factor. However, due to and leaf disease identification using image processing here
population and corresponding environmental issues such as Our working step process is three modules.
climate changing, pollution causes many problems to   Robot modules
farming such as disease & pests attacks.   Camera
In order this problems makes lot economic loss to the  Arduino Uno
people and government. And it’s also affect food
productivity causes to food shortage. Hence improve the
food productivity & sustainability there are lots of schemes
was introduced by the government. One of the thesis is
avoid the pest attacks and identify the disease affected leafs
before its spread whole cultivation. Image processing is one
of the technologies which can be used to identify the leaf
disease based on processing the images which were
captured by the cameras. Also to avoid human interference,
image processing module embedded with robotics where it
can be work as automated.
Detection of diseases in the plant is utmost need for farmers
and agricultural experts. In most of the plants the disease
inception takes place on plant leaves. The main aim of the

775
978-1-5386-9533-3/19/$31.00 ©2019 IEEE
2019 5th International Conference on Advanced Computing & Communication Systems (ICACCS 2019)

The design of Automatic Detection and order of leaf A. CAMERA


infection First, the continuous pictures of different leaves
are obtained utilizing a camera. Through Image Processing
the features of the image are extracted and the output is
given to Microcontroller. Then various image-processing
methods are applied to the acquired images to getting useful
features that are important for next analysis process. Feature
1) IMAGE ACQUISITION
extraction process will consists actual disease detection
The reason for this progression is to get the picture of an
from an image by comparing the image with non-defected
images. After that, numerous systematic procedures or entire plant or its organs with the goal that investigation
strategies are completed to characterize the pictures as per towards features can be performed.
the specific issue close by. 2) DIGITAL IMAGE PROCESSING
The proposed working process is leaf disease identification Digital image processing is the use of computer vision
using automation robot. In this project, robot module algorithms to perform image processing on digital images.
moving one plant to another plant through the guidelines of There lots statistical & mathematical methods used as
the Microcontroller. At that point as indicated by the illness, algorithms for the input image. There is wide range of
the sprayer will shower the bug spray to the leaves of the option to choose the algorithm. The computer vision
plant. algorithms explore a different kinds such enhancement,
color conversion, transforming techniques to perform. An
image is nothing more than a two-dimensional array of
numbers (or pixels) ranging between 0 and 255. It is defined
by the mathematical function f(x, y) where x and y are the
two co-ordinates horizontally and vertically. The value of
f(x, y) at any point is giving the pixel value at that point of
an image.
3) FEATURE EXTRACTION AND DESCRIPTION
There are lot of pattern algorithms, descriptor algorithms
and distance algorithms are available in openCV library
module. Feature extraction is applied the input image after
completing the pre-processing, colour conversion and
denoising. Where these techniques are used to improve the
Classification & Segmentation of input image.
4) CLASSIFICATION
Machine learning is niche technology which is dominantly
used in all applications in all industries. Machine learning
algorithms are nothing but statistical algorithms are
performing some mathematical operations in order to
specify the image class or category to apply classification.
In feature extraction the image features are converted into
vectors in order to specify the class type. With the input
feature vectors are matching or mapping with the train
Fig: Hardware Setup of Robot features then it being classified.
The leaf disease is identified by the image processing B. ARDUINO
technique using Python OpenCV library module. OpenCV
There are so many open source IDE platforms available for
means Open Computer Vision is created by Intel for CV
embedding applications and Arduino is one of the best IDE
applications. The 4GB RAM & 300GB ROM with Intel i3
to writing programs or creating projects in simple way . It
processor powered PC is used for CV processing. The next
includes a code editor with features such as syntax
is the Spray Fertilizer Control System which connected to
highlighting, brace matching, and automatic indentation.
PC through UART communication protocol.
Arduino version is used here for interfacing purpose. It is
First the CV module Capture image and process the image
also capable of compiling and uploading programs to the
to identify the leaf disease. After identification the disease
board with a single click. Arduino programs are written in C
the robot module operate the fertilizer to spray on leaf or C++. The board is equipped with 6 analog inputs and 14
otherwise the robot module moves to next leaf and digital input output pins. It acts as a communication
continues the process from beginning.
interface which can include USB interface used for

776
2019 5th International Conference on Advanced Computing & Communication Systems (ICACCS 2019)

programming. It consists of power jack, reset button, flash be identified at the severe stages only. The main objective of
memory of 32 kb, SRAM and EEPROM, clock speed the paper is to automatically detect and cure the plant
(16MHz) and it operates on 7 to 12 V. disease by providing medicine through IoT. In agricultural
field, the plant disease plays a vital role to cause loss
economically. The plant disease is identified by image
processing using the concept of histogram of gradients. The
histogram of gradients concept is used to zoom the image
and identified the affected part with more accuracy. Then
the severity of the disease is identified by comparing value
with the trained dataset and provides medicine accordingly.
In this proposed system the detection and curing of plant
disease will be done automatically. The medicine will be
providing through the ARDUINO microcontroller. The
microcontroller is serially connected with the system. The
proposed system will reduce the manual work and used to
increase the yield by identifying the disease in earlier stage.
Hence saving the loss and helps in agricultural field
efficiently.
Fig: Arduino
IV. REFERENCES
C. NODE-MCU
1] AakankshaRastogi, RitikaArora, Shanu Sharma, “Leaf
Node-MCU is an open-source mini-size micro-controller kit Disease Detection and Grading using Computer Vision
and firmware that helps you to prototype or build IoT Technology &Fuzzy Logic”
product. It includes firmware which runs on the ESP8266 [2] Haiguang Wang, Guanlin Li, Zhanhong Ma,
WiFi SoC from Espressif Systems, and hardware which is Xiaolong Li, “Image Recognition of Plant Diseases Based
based on the ESP-12 module. The firmware uses the Lua on Backpropagation Networks”
scripting language. It is based on the eLua project, and built [3] SaiKirthi Pilli1, BharathirajaNallathambi,
on the Espressif Non-OS SDK for ESP8266. The Node- Smith Jessy George, VivekDiwanji , “eAGROBOT- A
Robot for Early Crop Disease Detection using Image
MCU module is programmed by Lua programming and it’s
Processing”
similar to node.js programming. This Node-MCU is an [4] Fritz Brugger, “Mobile Applications in
asynchronous and event-driven. Many functions, therefore, Agriculture”, Syngenta Foundation, Basel,Swit zerland,
have parameters for call-back functions. Node MCU 2011
developers to create Lua applications with up to 256Kb Lua [5] Pierre SibiryTraoré, “The view from above” in ICT
code and read-only constants executing out of flash. The Update, a remote sensing scientist and GIS head at the
entire RAM is available for read-write data. (ICRISAT), 23 February 2010
[6] Lilienthal H, Ponomarev M, Schnug E 2004
Application of LASSIE to improve agricultural field
experimentation. LandbauforschVölkenrode 54(1):21-26
Online. Available:
http://literatur.vti.bund.de/digbib_extern/bitv/zi0 32847.pdf

Fig: Node Mcu


III. CONCLUSION
The smart agriculture using IOT has been experimentally
proven to work satisfactorily by monitoring the leaf diseases
in the agricultural field. Mostly identification and curing of
plant disease will be done manually. Then also disease will

777

S-ar putea să vă placă și