The document describes a proposed method for classifying rice leaf diseases using convolutional neural networks (CNNs). The method uses transfer learning with a DenseNet-201 model to classify images of rice leaves with different diseases. The model was trained on 1509 images and tested on 647 separate images, achieving an accuracy of 92.46%. Transfer learning helped improve performance given the small dataset size. Future work could involve collecting more images to further increase accuracy and applying other deep learning models for comparison.

1. Artificial Intelligence-Based Diseases
Classification For Rice Leaf Using CNN
Submited By
Vivek A
Aslin C
Ratheesh R
Guided by
Dr. Misba M

2. Abstract
 Rice is one of the maximum critical plants in India and is liable to diverse
illnesses at some point of extraordinary tiers of cultivation.
 It could be very tough for farmers with confined understanding to as it
should be perceiving those illnesses manually.
 Recent traits in deep gaining knowledge of have proven that automated
picture reputation structures the use of convolutional neural network (CNN)
fashions are very beneficial for such problems.
 The proposed CNN structure is primarily based totally on VGG-sixteen and
is educated and examined the use of paddy subject and net datasets.
 The accuracy of the proposed version is 92.46%.

3. Existing System
 A lot of research was completed using traditional classifiers but the
effects are relying at the characteristic desire techniques and photo
preprocessing is a high step.
 Therefore, CNN has attracted multiple researchers to take advantage
of immoderate reputation accuracy.

4. Disadvantages
 The classifier is transfer getting to know based definitely using Alex
Net.
 Training the above shape an accuracy of 91.23% is achieved but it can
maximum efficiently anticipate whether or not or now no longer plant
is diseased or now no longer.

5. Proposed Method
 In proposed system, we advise a Deep Learning generation that
automatically apprehends pics using Convolution Neural Network
(CNN) models can be very beneficial in such problems.
 By using the ones techniques, we are able to results easily discover
and select out the diseases.
 Our proposed approach we used Dense Net - 201 Model. This will
provide more accuracy

6. Advantages
 Predict more diseases
 High Accuracy Score
 Better Performance

7. Title Name Year Abstract Drawbacks
PLANT LEAF
DISEASE
ANALYSIS
USING
IMAGE
PROCESSING
TECHNIQUE
WITH
MODIFIED
SVM-CS
CLASSIFIER
T. Gupta 2017 This paper is mainly
developed to identify and
calculate the correctness of
pest infected area in leaf
images.
It takes time to
generate new
models.
SVM
CLASSIFIER
BASED
GRAPE LEAF
DISEASE
P. B. Padol
and A. A.
Yadav
2016 First the diseased region is
found using segmentation by
K-means clustering, then
both color and texture
features are extracted.
It will achieved
91.23%
accuracy.
Literature Survey

8. Literature Survey
Crop diseases have become a common part of the agricultural field and
with the growth of the agricultural field, these crop diseases are also increasing
day by day. Rice crop is one of the main crop and its plantation has spread in
almost every region of India and many parts of the globe also. Rice diseases are
very common and in recent decades various machine learning techniques have
been introduced to detect those diseases. In this paper, we have conducted a
survey study on eight major rice diseases namely bacterial leaf blight, false
smut, rice hispa, blast, stemborer, sheath blight, brown spot, brown planthopper,
and work conducted on them using CNNs technique. The paper is divided into
two major parts, first is the survey methodology followed for conducting the
work and second is state of the art used for rice disease detection (RDD) using
CNNs technique.
-Rishabh Sharma; Vinay Kukreja; Virender kadyan

9. Literature Survey
Rice is one of the maximum critical plants in India and is liable to diverse
illnesses at some point of extraordinary tiers of cultivation. It could be very tough for
farmers with confined understanding to as it should be perceiving those illnesses
manually. Recent traits in deep gaining knowledge of have proven that automated
picture reputation structures the use of convolutional neural network (CNN) fashions
are very beneficialfor such problems. Since rice leaf ailment picture datasets aren't
quite simply available, we created our very own small dataset. Therefore, I advanced
a deep gaining knowledge of version the use of switch gaining knowledge of. The
proposed CNN structure is primarily based totally on VGG-sixteen and is educated
and examined the use of paddy subject and net datasets. The accuracy of the
proposed version is 92.46%. Index Terms – Convolutional Neural Networks, Deep
Learning, Fine Tuning, Rice Leaf Disease, Transfer Learning.
-Paidi Haritha, Dr. R. Maruthamuthu M.C.A., Ph.D.

10. Literature Survey
Rice is one of the major cultivated crops in India which is affected by
various diseases at various stages of its cultivation. It is very difficult for the
farmers to manually identify these diseases accurately with their limited
knowledge. Recent developments in Deep Learning show that Automatic Image
Recognition systems using Convolutional Neural Network (CNN) models can
be very beneficial in such problems. Since rice leaf disease image dataset is not
easily available, we have created our own dataset which is small in size hence
we have used Transfer Learning to develop our deep learning model. The
proposed CNN architecture is based on VGG-16 and is trained and tested on the
dataset collected from rice fields and the internet. The accuracy of the proposed
model is 92.46%. Index Terms —Convolutional Neural Network, Deep
Learning, Fine-tuning, Rice leaf diseases, Transfer learning.
-Shreya Ghosal, Kamal Sarkar

11. Software requirement
 Environment - Jupyter Notebook
 Front End - Html,Css,Bootstrap,Flask
 Back End - Python
 Modules - Numpy,Pandas,Scikit-Learn

12. System Architecture

13. Modules
 Image Acquisition
 Image Preprocessing and Augmentation
 CNN Model Training
 Justification for the Chosen Model

14. 1.Image Acquisition
 The pictures are collected from the cultivation fields similarly to from
internet. As referred to within side the dataset description, statistics
encompass 4 schooling mainly Leaf Blast, Leaf Blight, Brown Spot
and healthful plant pictures
Modules Description

15. 2.Image Preprocessing and Augmentation
 The images amassed are resized to 224*224 pixel and a number of
augmentation techniques like zoom, rotation, horizontal and vertical
shift are achieved using Image Data Generator in Keras to generate
new images.

16. 3.CNN Model Training
 The picture information set is loaded for the education and checking
out.
 The elegance labels and the corresponding photos are saved in
respective arrays for education.
 70 percentage of information is used for education and 30 percentage
of information is used for checking out the usage of teach take a look
at cut up feature.
 The 70-percentage information is similarly cut up and 20% of its miles
used for validation.

17. 4.Justification for the Chosen Model
 Transfer getting to know refers back to the state of affairs wherein
what has been discovered in a single placing is exploited to enhance
generalization in some other placing.
 Transfer getting to know has the advantage of reducing the education
time for a neural community version and as a consequence could be
very beneficial considering the fact that maximum real-global troubles
normally do now no longer have hundreds of thousands of categorized
information factors to teach such complicated models.

18. Dense Net - 201
 Recent work has shown that convolutional networks can be
substantially deeper, more accurate, and efficient to train if they
contain shorter connections between layers close to the input and those
close to the output. In this paper, we embrace this observation and
introduce the Dense Convolutional Network (DenseNet), which
connects each layer to every other layer in a feed-forward fashion.
 We evaluate our proposed architecture on four highly competitive
object recognition benchmark tasks.
Algorithms

19. Screen Shots

20. Screen Shots

21. Screen Shots

22. Screen Shots

23. Conclusion
 In this paper we've got proposed a deep gaining knowledge of
structure with education on 1509 pictures of rice leaves and trying out
on one of a kind 647 pictures and that successfully classifies 92.46%
of the take a look at pictures.
 Transfer Learning the use of fine-tuning the predefined VGGNet has
significantly advanced the overall performance of the version which in
any other case did now no longer produce high-quality consequences
on such small dataset.

24. Future Enhancement
 We would really like to accumulate more pics from agricultural fields
and Agricultural Research institutes simply so we are able to decorate
the accuracy further.
 We would really like to characteristic cross-validation approach in
future a great manner to validate our consequences.
 We could additionally like to apply higher deep getting to know
fashions and different state-of the artwork works and examine it with
the outcomes obtained.
 The evolved version may be utilized in destiny to locate different plant
leaf diseases, that are crucial vegetation in India

25. Thank
You