CROP DISEASE DETECTION

RAGHU ENGINEERING COLLEGE

Autonomous
DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING

Presented By
G. Swarna (203J1A0566)
N. L. Sunandhitha (203J1A05C0) Under the guidance of
K. Haritha Kumari (203J1A0587)
Mrs.M.S.R.Pavani
M. Appala Naidu (213J5A0510)
 Abstract

• Nowadays many of the farmers and agro help center use the different new
technology to enhance the agriculture production. Plants have become important
source of energy. There are several diseases that affect plants with the potential to
cause economic and social losses. Disease detection along with proper monitoring of
soil moisture, salinity and pH is very important for high yield and increasing the
production. Thus, this paper provides the best method for detection of plant diseases
using image processing and alerting about the disease caused by sending email. It
also discusses about how the moisture of the soil, pH and salinity of water is
monitored and controlled using soil moisture sensor, pH sensor and salinity sensor.

• Link for base paper: https://www.ijrti.org/papers/IJRTI1805054.pdf

 Introduction
• Optimal agriculture is vital for global development, as human survival hinges on
productive farming. Challenges include early crop diseases, inadequate soil
monitoring, and water management. In India, diverse crops face pathogen-
related issues, impacting yield. Leaf discolorations signal diseases.

• Traditional observation methods were inefficient. Modern approaches require

labs and expertise, limiting remote adoption. This paper introduces an image
processing-based solution for early disease detection and comprehensive plant
monitoring. The methodology covers disease identification through images,
while the second part explains soil moisture, water, and pH management. This
innovative approach addresses agricultural limitations, enhancing productivity .
 Existing System

• The existing crop disease detection system relies primarily on manual visual
inspections by farmers and experts. This labor-intensive process often leads to delayed
disease identification and subsequent treatment, resulting in substantial crop losses.
Limited by human subjectivity, the current system lacks the precision and speed
required for effective disease management. It also lacks real-time monitoring
capabilities, making it challenging to provide timely interventions. In contrast, the
proposed system aims to overcome these limitations through automated image
analysis, harnessing the power of machine learning to enable early disease detection
and swift decision-making for improved agricultural productivity.
 Proposed system

• The proposed crop disease detection system integrates advanced machine learning
techniques with real-time image analysis to revolutionize agricultural disease
management. Leveraging a diverse dataset of healthy and diseased crop images, the
system employs convolutional neural networks for accurate disease identification.
Through a user-friendly mobile app, farmers can swiftly capture and upload images of
their crops, receiving instant diagnoses and actionable recommendations. The system's
scalability and adaptability ensure its continuous enhancement to encompass new
crops and diseases, while stringent ethical considerations underscore its positive
impact on sustainable farming practices.
 Software specifications

Software system configuration:-

 Operating System - Windows XP

 Coding Language - Python

 Front End - J2EE

 Back End - MySQL

6
 Modules

• Image acquisition
• Image preprocessing
• Image segmentation
• Feature extraction
• Statistical analysis
• Classification based on a classifier

7
 Modules

Image acquisition

• Image acquisition refers to the process of capturing visual

information from the real world and converting it into a digital
format that can be stored, processed, and analyzed using
computers and software. This is a fundamental step in various
fields such as photography, remote sensing, medical imaging,
computer vision, and more.

8
 Modules
Image preprocessing

• Image preprocessing in crop disease detection involves preparing raw

images for analysis by enhancing their quality and extracting relevant
features. This improves the accuracy of disease identification
algorithms. Common preprocessing steps include noise reduction,
resizing, and adjusting brightness and contrast. Additionally,
techniques like image segmentation isolate the crop from the
background, making it easier to identify disease symptoms. These
preprocessing steps ensure that the subsequent disease detection
algorithm works effectively and provides accurate results.
9
 Modules

Image segmentation

• Image segmentation in crop disease detection is a process that

involves dividing an image into distinct regions, making it easier to
analyze and identify specific areas affected by diseases. This
technique helps isolate and highlight the regions of interest within the
image, enabling accurate identification and quantification of disease
symptoms. By segmenting the image, the algorithm can focus on the
relevant portions and provide targeted insights for early disease
detection and precise treatment.

10
 Modules
Feature extraction

• Feature extraction is a critical step in crop disease detection using

machine learning. It involves transforming raw image data into
meaningful features that capture relevant patterns and characteristics of
the crops and their diseases. Techniques like Histogram of Oriented
Gradients (HOG), Local Binary Patterns (LBP), and deep learning-
based methods like Convolutional Neural Networks (CNNs) can be
employed for effective feature extraction. These features serve as input
for the disease classification model, aiding in accurate identification
and prediction of crop diseases.
11
 Modules
Statistical analysis

Statistical analysis of leaves involves applying various statistical techniques to data related to
leaf characteristics. These analyses can provide insights into plant biology, ecology,
environmental conditions, and more. Here are some specific statistical analyses that can be
performed on leaf-related data:

• Leaf Size Analysis: Calculate descriptive statistics such as mean, median, standard
deviation, and range of leaf sizes within a population.

• Leaf Shape Analysis: Extract shape features like aspect ratio, perimeter, and area from leaf
outlines.

• Leaf Venation Analysis: Quantify characteristics of leaf veins, such as vein density and
branching patterns.

• Leaf Color Analysis: Use color descriptors to analyze color features of leaves.
 Modules
Classification based on a classifier

• Data Collection: Gather labeled leaf images for different types.

• Feature Extraction: Extract key traits like shape, color, and texture.

• Training and Testing:Split data into training (70-80%) and testing (20-30%) sets.

• Choose Classifier: Pick algorithms like Decision Trees, SVM, or CNN.

• Model Training: Train classifier on training data to learn patterns.

• Model Evaluation: Test on testing data to measure accuracy.

• Refinement: Adjust parameters for better performance.

• Prediction: Use model to classify new, unseen leaves.

• Application: Identify plant species, detect diseases, etc.

• Insights: Gain insights into plant diversity and health.

Proposed approach

14
Use Case Diagram
 Conclusion

• The objective of this work is the crop disease detection, classification of leaf
diseases using image processing tools and also monitoring the important factors
such as soil moisture, water level and pH value which are of course the important
factors for high yield.

• All information about the disease and the important factors is sent to the farmer’s
mobile phone via mail.

• To increase the speed and accuracy of detection as well as classification of leaf

diseases we are using Raspberry pi module.

• One more important benefit of this system is that it also sends the image of the
affected crop or the leaf to the farmer, such that it becomes easier for the farmer to
take the precautionary measures.
17
 References
[1] Sanjay B. Patil , Dr. Shrikant K. Bodhe, “LEAF DISEASE SEVERITY
MEASUREMENT USING IMAGE PROCESSING”, International Journal of
Engineering and Technology Vol.3 (5), 2011.

[2] Prof. Sanjay B. Dhaygude, Mr. NitinP. Kumbhar, “Agricultural plant Leaf Disease
Detection Using Image Processing”, International Journal of Advanced Research in
Electrical, Electronics and Instrumentation Engineering, Vol. 2, Issue 1, January 2013.
[3] Murali Krishnan, Jabert.G, “Pest Control in Agricultural Plantations Using Image
Processing”, IOSR Journal of Electronics and Communication Engineering (IOSR-
JECE) Volume 6, Issue 4 May. - Jun. 2013.

[4] M. Reyalat ,Al Hiary , Bani Ahmad,Braik and Z. Rahamneh, “Fast and Accurate
Detection and Classification of Plant Diseases”.

[5] Smith J S., Camargo A., “An imageprocessing based algorithm to automatically
identify plant disease visual symptoms”.
[6] Sulaiman, N., Mustapha,“K-means clustering to improve the accuracy”.
[7] David Mount,Tapas kanungo, “An Efficient k-means Clustering Algorithm:
Analysis and Implementation”.
[8] Ford A., Robert, “Colour Space Conversions”,.
[9] C M Procopiuc, P. K. Agarwal, “Exact and Approximation Algorithms for
Clustering”