: It is an End to End deep learning project to classify
disease in plants .I have built a web application in this project that can take a
picture of the plant and tell the farmer if the plant has a disease or not.
1. Plant Disease Detection Using Machine
Learning
A project Submitted in partial fulfillment of the
requirements
for the award of the Degree of
MASTER OF COMPUTER APPLICATION
By
Guide: Prof. Manabendra Nath
Name of Student: KOUSHIK HAZRA
Roll Num: 12021010010006
Name of Student: DIBAS KUMAR SHYAMAL
Roll Num: 12021010010008
Name of Student: DEBODIP HAIT
Roll Num: 12021010010050
Name of Student: SUBHA CHATTERJEE
Roll Num:12021010010023
Name of Student: JINIYA MANDAL
Roll Num: 12021010010035
Name of Student: ADREJA SAHA
Roll Num: 12021010010030
3. INTRODUCTION
In this project, we have created a
convolutional neural network which will be able
to predict whether a plant is suffering from a
disease or not.
We used different layers and other
hyperperameters for building, training, and
testing this classification model.
We have used Tensorflow and Keras for this
project.
5. IMAGE ACQUISITION
The first step in plant disease
detection using machine learning
is to acquire images of plant
leaves that are both healthy and
diseased. Images can be
captured using cameras, smart
phones, or drones, and can be
taken in a controlled
environment, such as a
greenhouse, or in the field. It is
important to ensure that the
images are of high quality, with
good resolution and color
accuracy, to ensure accurate
disease detection.
6. DATASET PREPARATION
The Dataset collected from open source website “Kaggle”.
Corn (Maize) – Common Rust
Potato – Early Blight
Tomato – Bacterial Spot
7. IMAGE PREPROCESSING
Once the images are
acquired, they need to be
preprocessed to improve
their quality and make
them suitable for
analysis.
The goal of image
preprocessing is to make
the images more uniform
and to remove any
variability that could
interfere with disease
detection
For normalize our dataset
we will convert the
images into a numpy
Cropping
Filtering
Resizing
Normalizatio
n
Imag
e
Preprocess
ed
image
8. IMAGE SEGMENTATION
After preprocessing, the images are
segmented to separate the plant leaves
from the background.
Image segmentation involves dividing the
image into multiple regions, each of which
contains pixels with similar properties.
This allows the machine learning algorithm
to focus on the plant leaves and ignore the
background, which can improve disease
detection accuracy.
9. FEATURE EXTRACTION
Once the plant leaves are segmented, the next
step is to extract features from the image that
can be used to train the machine learning model.
These features might include color histograms,
texture features, or shape descriptors.
10. DATASET SPLITTING
The preprocessed images are labeled as
healthy or diseased and used to create a
dataset. The dataset is divided into training,
validation, and testing sets.
To split the dataset into testing and training
data. Here we have taken test size as 0.2 so my
data will be divided into 80% training and 20%
testing data.
11. MODEL TRAINING
Convolutional Neural Networks (CNNs) is chosen to train the
model and extracted features to recognize the visual symptoms
of plant diseases.
We have used different types of layers according to their
features namely
Conv_2d (It is used to create a convolutional kernel that is
convolved with the input layer to produce the output tensor)
max_pooling2d (It is a down sampling technique which takes
out the maximum value over the window defined by poolsize)
flatten (It flattens the input and creates a 1D output)
Dense (Dense layer produce the output as the dot product of
input and kernel).
13. TESTING AND VALIDATION
Fitting the model with the data and finding out
the accuracy at each epoch to see how our
model is learning.
The trained model is then tested on a separate
dataset of images to evaluate its accuracy in
detecting plant diseases.
15. FUTURE SCOPE OF
PROJECT
Our project have shown pretty good accuracy,
it can be implemented in real time mobile
applications and web services, so that formers
can identify diseases simply by taking photo of
suspected leaves of plants.
Other than plant leaf disease identification, it
can also be used for identification and
classification of nutrients deficiency of plant
leaves.
16. THE STEPS TAKEN TO SOLVE THE
PROBLEM
We started with loading the dataset into googlecolab
using Google drive and visualizing the images.
Normalizing is an important step when working with any
type of dataset. After that we created a CNN Model
which is further used for predicting the plant diseases
using the image supplied to model.
This model is highly beneficial as it can be used by
different agricultural firms and farmers to increase their
yield and stop wastage of crops due to disease.
CONCLUSION