MONITORING CROP HEALTH

Team Id:………………….

Ranu Pandey (2101641520114)

Krati itondia (2101641520081)
Muskan Shah (2101641520093)
Shreya Shukla (2101641520133)
Himanshi Singh (2101641520074)
INTRODUCTION :
 India is one of the nation where agriculture are allied sectors are major employment
sources.
Thus, an efficient monitoring system is required for continuous and long term plant health
monitoring.
Plant health monitoring is an essential in today’s world due to the climatic changes , which
affects the growth of the plants and their productivity.
The project is focusing on the main areas where the farmers today are facing problems for
the maximum cultivation and thus providing them as integrated system.
The proposed model presents a smart Agricultural System model divided into the following
phases:-
Crop Vegetation Information
Plant Disease Detection
Team_Id: 2
OBJECTIVES:
This project is to predict the maximum yield of the crops produced at
minimum cost.
Early detection and management of problems associated with crops yield
indicators can help increases yield and economic growth.
By influencing regional weather patterns, large scale meteorological
phenomena can have a significant impact on agricultural production.

Team_Id: 3
PROBLEMS:
• In general, the rainfall, temperature, pH is low and highly variable which
results .In uncertain crop yields. Besides its uncertainty, the distribution of
rainfall during the crop period is uneven, receiving high amount of rain,
when it is not needed and lack of it when crop needs it.
• Manures ,Fertilizers , and Biocides
• Lack of infrastructure

Team_Id: 4
SOLUTION:
• Inthis project we have developed a crop health monitoring system, which
will help a farmer to identify any visible disease in the crop leaf and help
him to
limit spread of the disease throughout the field.

The sensors installed in the field will continuously collect data about
physical parameters such as temperature, humidity of atmosphere, moisture
content in soil and rain detection.

The project also aims to alert farmers if any of the atmospheric parameters is
unfavorable for plant health and growth.

Team_Id: 5
FEASIBILITY:
• The main goal of agriculture app development is to optimize farming
processes. Since most features for smart farming are already available on
desktop computers and laptops, extending them to mobile devices is the next
logical step, allowing farmers to use these technologies wherever they are.

Team_Id: 6
TECHNICAL FEASILBILTY:
1) The core aim of the present study is to establish a model which will
efficiently predict a crop based on a set of features.

2) Machine learning algorithm use in this model.

3) In this we use PyCharm as IDE with a complete set of tool for python
development. It codes faster with more easier smart and configurable editor
with code completion snippet, code folding split window support.

Team_Id: 7
OPERATIONAL FEASIBILTY:
1) To collect data of ground parameters affecting the growth of the crop .

2) To create a solution for estimating the agricultural land vegetation.

3) Incorporate network communication between the system and cloud server

for proper transmitting effective information to the end users, i.e farmers.

Team_Id: 8
ECONOMICAL FEASIBILITY:
1) To increase profitable farm income.

2) Enhance quality of life for farm family and community.

3) Increase production for human food and fiber needs.

Team_Id: 9
LEGAL FEASIBILTY:

Team_Id: 10
Methodology:
It divided into three parts
1. Block Diagram
2. Mobile Applications
3. System Module
4 Software

Team_Id: 11
 Raw data
Block Diagram collection

Missing Pre-
Data Processing

Sampling

Training Split
Feature Selection Dataset
Dimensionality Pre-
Reduction Processing Test New
Learning Algorithm
Dataset DataPredictio
Final model n
Training evaluation
Hyperparameter
Optimization

Post-Processing

Final Classification/Regression
Model
Team_Id: 13
MOBILE APPLICATION:

Team_Id: 14
 MONITORING CROP
HEALTH

LOGIN ID:

PASSWORD:

LOGIN

Team_Id: 15
(A)SENSOR MODULE:
Temperature and humidity sensor (DHT11), Soil moisture sensor and rain sensor will collect data every 3 seconds.
Temperature, moisture values will be compared with their respective thresholds which are different according to the
plant.

If the threshold is crossed the farmer will get an alert message on his phone. To send the alert message we will use
GSM module. According to the reading of soil moisture sensor and
rain sensor irrigation pump will be actuated to turn on or off the water supply

Team_Id: 16
WORKING DIAGRAM:

Team_Id: 17
(B) SOFTWARE:-
Steps to follow

step 1:Import of libraries

step 2:Define Path to Image Dataset
step 3:Load and Normalize Dataset using ImageDataBunch.from_folder
step 4:Create transfer learning model using cnn_learner and resnet 34 model
step 5:Save The Model
step 6:Find the learning rate and plot the graph
step 7:Deploy Nodel Using Docker

Team_Id: 18
1) Creating Plant Disease Detection
Model:
Plant disease detection model was created using Fast.ai library.Firstly,
dataset was imported out of which 80% of dataset is being used as training and
rest for testing. Resnet-34 which is a pre trained model was used for deeper
learning. After the dataset, all the libraries were imported and the path for the
dataset was defined. To get the labels of all the folder names,
ImagedataBunchfrom_folder was used. After this the data was normalized to
imagenet parameters.

2) Converting Trained Model into Web

Application: The necessary packages were downloaded which
includes starlette as framework and Uvicorn as ASGI server. Next step was to
load the model which was trained earlier. After this, the classes are defined
which are supposed to be detected by our application..
Team_Id: 19
TOOLS And TECHNOLOGY:
The project is basically divided into two modules:

1. Hardware Requirements: The sensor module includes three sensors which are temperature and humidity sensor (DHT11), Soil moisture sensor and Rain detection module. A microcontroller to analyse the data from these sensors and interface sensors with a GSM module

2. Software Requirements: To create a web application for detection of plant diseases many different deep learning libraries have been used like Fastai which runs over Pytorch. The platform used to execute the code is Jupyter notebook and ResNet 34 model was used for the deeper training of the model and to gain high accuracy.

Team_Id: 20
HARDWARE REQUIREMENT:
1.Soil moisture sensor
2.Temperature and humidity sensor
3.Rain detection module
3.Micro Controller
4.GSM module

Team_Id: 21
SOFTWARE REQUIREMENTS:
1.Ardino IDE
2.Jupyter Notebook
3.Fast.ai
4.Crop disease Dataset
5.ResNet-34 model
6.Libraries for Deploying model

Team_Id: 22
TECHNOLOGY USED:
 HTTP
HTTP defines the precise manner in which web clients communicate with web servers.
 MY SQL
The name SQL stands for structural query language. SQL communicates with database
manager. The database
manager could be oracle, etc.
 HTML
Hyper text mark-up language is the standard language for creating documents for the
world wide web.
 CSS
CSS stands for cascading style sheet. CSS is used to design HTML tags. CSS is a
widely used language on the web.

Team_Id: 23
REFFERENCE:
• [1]
Neha Bhati, “Sense Pi: Plant health monitoring and local database
connectivity using Raspberry Pi ”, International Journal for Research in Applied
Science & Engineering Technology (IJRASET),vol. 3,2015

• [2]Bhavana Patil, “Plant monitoring using Image Processing Pi & IoT” ,

International Research Journal of Engineering and Technology(IRJET), vol.
4,2017.

• [3]M.OZYAVUZ,B.C.BILGILI,A.SALICI, “Determination of vegetation

changes with NDVI Method”, Journal of Environment Protection and Ecology,
vol.16,No.1,pp.264-273,2015.
Team_Id: 24
Team_Id: 25
Team_Id: 26
Team_Id: 27
Team_Id: 28
Team_Id: 29
Team_Id: 30