DISEASE DETECTION

IN
PADDY USING
MACHINE LEARNING

TEAM MEMBER'S:
BANDI RAKSHITHA : 21C31A0576
AARKATLA NAVYA : 21C31A05B3
G.ANISH : 21C31A0588
AZMATH AFREEN : 21C31A05D4
 Abstract​  Gaps identified
 Motivation and Background  Architecture
 Objectives  Proposed Solution
 Introduction  Result
 Problem Statement  Conclusion
 ​Literature Survey
 Existing Solution
 Disadvantages
ABSTRACT

This work represents a machine learning model .We used python language and ensembled learning
techniques in order to predict the disease in paddy. The proposed system will integrate the data
obtained from a dataset where around 1500 images are clicked and set up to dataset. .By using
machine learning algorithms we can identify the diseases and classify them into different types that
gives farmer a better idea about the crop. This provides farmer information about their crop and what
disease his crop has, such that he can use appropriate fertilizer or pesticide to get rid of that disease.
The diseases in plant are due to many reasons like soil, excess rains, weather and climate change.
Paddy is the most grown crop in our country and agriculture is a leading sector in our country .So, it
would be definitely for farmers in real-time.
MOTIVATION AND BACKGROUND

We know Agriculture system is supported by farmers. Agriculture is the major sector and many people
live on that sector. It contributes highly to the economy of our country. So, to assist farmers in more
enhanced way to train the newer farmers to know about the disease in plants and which crop should be
cultivated in a particular condition, such that they get a clarity on cultivating crops.
OBJECTIVES

• 1.Early Detection: Detecting diseases in rice plants at an early stage can help
prevent the spread of diseases and minimize crop damage.
• 2.Improved Crop Management: By accurately identifying diseases, farmers can
implement targeted management strategies such as applying appropriate
pesticides or adjusting irrigation and fertilization practices.
• 3.Increased Crop Yield: Minimizing the impact of diseases on rice crops can lead
to higher yields and better overall productivity.
• 4.Reduced Economic Losses: Timely detection and management of diseases can
reduce economic losses for farmers by preventing extensive damage to crops.
INTRODUCTION

• Paddy, one of the most crucial cereal crops globally, is particularly

susceptible to a multitude of diseases that can decimate yields if left
unchecked. Traditional methods of disease detection rely heavily on
manual inspection, which is not only time-consuming but also often
prone to human error. In recent years, the integration of machine
learning (ML) techniques into agricultural practices has sparked a
revolution in disease detection and crop management. Harnessing the
power of artificial intelligence, specifically ML algorithms, offers a
promising solution to mitigate the impact of diseases on paddy
cultivation.
PROBLEM STATEMENT

• Agriculture is the major sector in our country.And we proudly say that farmers are the backbone of
our country.As we can see paddy is the most common and grown crop in India.It could be dealt
with the various diseases and some may not know the variety of the disease.Without knowing the
disease we cannot know the appropriate pesticide or fertilizer to use for that particular disease.So,to
classify the disease and solve these problems we can use the modern technology like using machine
learning and Artificial intelligence (AI).
LITERATURE SURVEY

• V.Vanitha proposed “Rice disease detection using deep learning” which proposes an automatic disease detection using deep neural
network. The developed model is capable of detecting 3 different disease of paddy and also detect the healthy leaf image.The
dataset was trained with three CNN architectures and achieved a high accuracy of 99.53% [1].

• Anuradha badge proposed “Crop disease detection using Machine learning: Indian Agriculture” how diseases affecting the less
yield and how machine learning technique will help us to detect the disease and help the farmers to take necessary action. They
used canny’s edge detection algorithm for the

• efficient detection of crop disease by taking image of crop .They take wheat crop for their research paper [2].

• Prabakaran,G et al., has presented the “Fuzzy decision support system for improving the crop productivity and efficient use of
fertilizers' ' states that the process of reducing the fertilizer consumption and improving the crop productivity using the fuzzy logic
systems.[3]
EXISTING SOLUTIONS

• Visual Inspection: Farmers and agricultural workers visually inspect crops for signs of disease.
This method involves observing the appearance of plants, including changes in color, shape,
size, and texture of leaves, stems, and other plant parts. Visual inspection requires knowledge
and experience to recognize disease symptoms accurately.
• Symptom-based Identification Guides: Agricultural extension services often provide guides
or manuals containing descriptions and images of common crop diseases. Farmers can compare
the symptoms observed on their crops with those depicted in the guides to identify potential
diseases.
• Field Surveys: Agricultural experts conduct field surveys to assess the health of crops in a
particular area. These surveys may involve systematic sampling of plants, collection of data on
disease prevalence and severity, and visual inspection of multiple fields to identify patterns and
trends in disease occurrence.
DISADVANTAGES OF EXISTING
METHODOLOGIES
• Subjectivity: Visual inspection and symptom-based identification rely on
the subjective interpretation of disease symptoms by farmers or
agricultural experts. This subjectivity can lead to inconsistencies in
disease diagnosis, especially among individuals with varying levels of
expertise.
• Field surveys and visual inspection are limited by the physical
accessibility of agricultural fields. Certain areas, such as densely
vegetated or geographically remote regions, may be challenging to
survey effectively, resulting in gaps in disease monitoring and
surveillance.
PROPOSED SOLUTION

• This project presents a rice leaf disease detection system using machine learning
approaches. Three of the most common rice plant diseases namely leaf smut, bacterial
leaf blight and brown spot diseases are detected in this work. Clear images of affected
rice leaves with white background were used as the input. After necessary pre-processing,
the dataset was trained on with a range of different machine learning algorithms and we
are using linear support vector machine(svm).
Start (Installation & Imports):
This section represents the initial steps of your code where you start by installing required
packages and importing necessary libraries.Here, you perform tasks such as installing packages like
OpenCV, numpy, mahotas, scikit-learn, etc., using pip, and importing libraries/modules like cv2,
numpy, os, glob, mahotas, etc., into your Python environment.

Image Processing and Feature Extraction:

This section encompasses the processing of images and extraction of features from them.Tasks in
this section include reading images, resizing them to a standard size, converting them to grayscale
(if necessary), and extracting relevant features using methods like Haralick texture
extraction.Image processing techniques like resizing and color space conversion are applied here to
prepare the images for feature extraction.

Model Training and Evaluation:

Here, you train your machine learning model using the extracted features and corresponding
labels.This section involves creating a classifier (in this case, a Linear Support Vector Machine) and
fitting it to the training data.Evaluation of the model's performance is also conducted, typically
using techniques like cross-validation to assess its accuracy and generalization ability.
Model Deployment and Prediction:
This section involves deploying the trained model to make predictions on new, unseen data.After
training, the model is saved for future use.In deployment, the saved model is loaded, and
predictions are made on test or real-world data using the learned patterns from the training phase.
RESULT AND ANALYSIS

we are getting output like this:

CONCLUSION

• In conclusion, our study demonstrates the potential of machine learning techniques in

accurately detecting diseases in paddy crops. Through the utilization of advanced
algorithms and feature engineering, we achieved promising results in terms of disease
identification and classification accuracy. Our findings underscore the significance of
timely disease detection in agricultural practices, particularly in the context of paddy
cultivation where diseases can significantly impact crop yield and quality.