You are on page 1of 19

Savitribai Phule Pune University, Pune

Amrutvahini College Of Engineering, Sangamner


Department of IT Engineering

A Presentation
on
“Crop Disease and Pest Detection Using CNN”
Presented By

Ms. Khatode Pratipada Madhukar


Ms. Mhase Mansi Ganesh
Ms.Urhe Tejal sharad
Ms.Dagale Dipali chandrkant
Under the Guidance of

Dr. B.S.Borkar
CONTENT

 Problem Statement
 Motivation
 Purpose
 Scope
 Objectives
 List of required software, Test Environment , Cost
 Architecture
 Conclusion
 Future Scope
 Reference

02/25/24 Crop disease and pest prediction using CNN 2


Problem Statement

Crop Disease and Pest Detection


Using CNN

02/25/24 Crop disease and pest prediction using CNN 3


• Scope of topic:

 The scope of the Crop Disease & Pest Detection project involves the
development of a computer vision system based on Convolutional Neural
Networks (CNNs) to detect and diagnose diseases and pests in agricultural
crops.
 This project aims to cover a range of crops and their associated diseases and
pests.
 It may also encompass multiple stages of crop growth, from planting to
harvest.

02/25/24 Crop disease and pest prediction using CNN 4


Objectives:
• Develop a robust CNN-based model for crop disease and pest detection. Create a
dataset of labeled images covering various crops, diseases, and pests.
• Implement a user-friendly interface for farmers or agricultural experts to interact
with the system.
• Provide real-time or near-real-time detection capabilities to aid in timely
intervention.
• Enhance accuracy and reduce false positives to increase the reliability of
detection.
• Explore the possibility of integrating the system with mobile applications or
drones for field-level monitoring.

02/25/24 Crop disease and pest prediction using CNN 5


Motivation:
The motivation behind this project is to address several key challenges in agriculture:
Crop Yield Loss:
Diseases and pests can significantly reduce crop yields, leading to economic
losses for farmers.
Food Security:
Ensuring healthy crop growth is crucial for global food security.
Sustainability:
Early detection can minimize the need for chemical pesticides and promote
sustainable farming practice.
Accessibility:
Providing a user-friendly tool for farmers to identify issues in their crops,
especially in remote areas.

02/25/24 Crop disease and pest prediction using CNN 6


List OF Required:
1.Hardware:
A. Computer/Server:
High-performance computer or server with a powerful CPU and GPU (Graphics
Processing Unit) to train and run deep learning models efficiently.
B. GPUs:
Modern GPUs are essential for training deep CNNs effectively. NVIDIA GPUs,
such as the GeForce or Tesla series, are commonly used for deep learning tasks.
C. Data Storage:
Sufficient storage for storing large datasets of images, model checkpoints, and
intermediate results.
D. Camera or Image Acquisition Devices:
If you're capturing images from the field, you may need cameras or image
acquisition devices. These could be standard digital cameras or specialized sensors.

02/25/24 Crop disease and pest prediction using CNN 7


E. IoT Devices (Optional):
If you're implementing real-time monitoring in agricultural fields, you may require IoT
devices like cameras, sensors, or drones to capture images and data.
F. Mobile Devices (for Interface):
If you plan to provide a mobile interface for farmers, you'll need mobile devices for
development and testing.
G. Networking Equipment:
Reliable network connectivity to transfer data between devices and potentially for
cloud-based processing.
H. Sensors (Optional):
Depending on the specific requirements of your project, you might need
environmental sensors to collect additional data about temperature, humidity, soil
conditions etc.
I. Power Supply:
Uninterrupted power supply, especially for field-based installations to ensure
continuous operation.

02/25/24 Crop disease and pest prediction using CNN 8


J. Development Boards (Optional):
Raspberry Pi or similar boards if you plan to deploy low-power, edge-based solutions
for real-time monitoring in remote areas.
K. Mobile Devices for Testing:
Smartphones or tablets for testing the user interface of mobile applications.
L. Drone (Optional):
If you intend to use drones for aerial image capture, you'll need the drone hardware
and related equipment’s.

02/25/24 Crop disease and pest prediction using CNN 9


2.Software:
A. Deep Learning Framework:
Choose a deep learning framework like TensorFlow, PyTorch, or Keras that supports
CNNs. These frameworks provide pre-built layers and tools for building and training
neural networks.
B. Python:
You'll need Python as the programming language for implementing your CNN model
and handling data.
C. Data Preparation Tools:
Libraries like NumPy, Pandas, and OpenCV can be used for data preprocessing and
augmentation.
D. GPU Drivers and CUDA:
If you're using a GPU, you'll need to install the appropriate GPU drivers and CUDA
(Compute Unified Device Architecture) toolkit for GPU acceleration.
E. Jupyter Notebook or IDE:
A development environment like Jupyter Notebook or a code editor (e.g., Visual Studio
Code, PyCharm) to write and run your code.
02/25/24 Crop disease and pest prediction using CNN 10
F. Data:
Access to a dataset of crop images with labels for diseases and pests. Ensure your
data is properly organized into training and testing sets.
G. Pre-trained Models:
You may benefit from pre-trained CNN models like VGG16, ResNet, or Inception,
which can be fine-tuned for your specific task.
Dependencies: Install any required Python packages and libraries, such as
TensorFlow or PyTorch, using package managers like pip or conda.
H. Virtual Environment:
It's a good practice to create a virtual environment to manage your project's
dependencies and avoid conflicts.
I. Data Annotation Tools:
If your dataset isn't already labeled, you may need annotation tools for labeling
images with disease and pest information.
J. Version Control:
Consider using a version control system like Git to track changes in your codebase.

02/25/24 Crop disease and pest prediction using CNN 11


K. Documentation and Reporting:
Tools like Jupyter Notebook, Markdown, or LaTeX for documenting your
project, experimental results, and findings.
L. Testing Framework:
Implement a testing framework to evaluate the model's performance, such as
accuracy, precision, recall, and F1-score.
M. Deployment Environment (Optional):
If you plan to deploy your model for real-time predictions, you'll need an
appropriate server or cloud environment.
N. Web Interface (Optional):
If you want to create a user-friendly interface for users to interact with your
model, you may need web development tools and frameworks.

02/25/24 Crop disease and pest prediction using CNN 12


Architecture:

02/25/24 Crop disease and pest prediction using CNN 13


02/25/24 Crop disease and pest prediction using CNN 14
Conclusion:
• Significance of Crop Disease and Pest Identification:
– Crucial for modern agriculture.
– Integral for effective management strategies.
• Role of CNN Technology:
– Utilizes technologies like Convolutional Neural Networks (CNN).
– Enables accurate and timely detection of diseases and pests.
• Benefits to Farmers:
– Minimizes crop losses.
– Optimizes resource utilization.
– Promotes sustainable farming practices.
• Advanced Image Recognition Techniques:
– CNN facilitates automated and efficient analysis of crop images.
– Enables early detection of diseases and pests..

02/25/24 Crop disease and pest prediction using CNN 15


• Prompt Action and Interventions:
– Early detection allows farmers to take prompt action.
– Implementation of targeted treatments to mitigate threats.
• Informed Decision-Making:
– Identifying specific diseases or pests.
– Informed decisions on control measures, including pesticides or biological
agents.
• Contribution to Research and Development:
– Provides valuable insights and data.
– Facilitates large-scale data collection and analysis.
• Identifying Patterns and Trends:
– Helps identify patterns, trends, and potential outbreaks.
– Aids in the development of more effective prevention and control strategies

02/25/24 Crop disease and pest prediction using CNN 16


Future Scope:
• The future scope of crop disease and pest identification looks promising.
Advancements in technology, such as artificial intelligence and machine
learning, can enhance the accuracy and efficiency of identification systems.
This can lead to early detection, precise targeting of treatments, and improved
crop management practices

02/25/24 Crop disease and pest prediction using CNN 17


Reference:
• 1. J. Liu, X. Wang, “Plant diseases and pests detection based on deep learning:
a review,” vol. 17, pp. 1-18, 2021. Retrieved from
https://plantmethods.biomedcentral.com/arti 021-00722-9.
• 2. S. Magalha ˜es, et al. ”Evaluating the single-shot multibox detector and
YOLO deep learning models for the detection of tomatoes in a greenhouse.”
Sensors, vol. 21, 3569, 2021.
• 3. Y. Qiu, et al. “Pose-guided matching based on deep learning for assessing
quality of action on rehabilitation training.” Biomedical Signal Processing and
Control, vol. 72, 2022.
• 4. G. Dhiman, et al. “A Novel Machine-Learning-Based Hybrid CNN Model
for Tumor Identification in Medical Image Processing.” Sustainability. 2022;
14: 1447.” 2022.

02/25/24 Crop disease and pest prediction using CNN 18


THANK YOU

02/25/24 Crop disease and pest prediction using CNN 19

You might also like