├── documentation
    ├── project_proposal.pdf
    ├── project_report.pdf
    └── assets
    │   └── images
    │       ├── home_page.png
    │       ├── compare_page.png
    │       ├── predict_page.png
    │       └── results_page.png
├── web_application
    ├── static
    │   ├── bg.jpg
    │   ├── comparison_graph.png
    │   ├── temp
    │   │   └── UNADJUSTEDNONRAW_thumb_1d.jpg
    │   └── style.css
    ├── templates
    │   ├── bg.jpg
    │   ├── prediction.html
    │   ├── comapre.html
    │   ├── result.html
    │   └── index.html
    └── main.py
├── data
    └── README.md
├── LICENSE
├── requirements.txt
└── README.md


/documentation/project_proposal.pdf:
--------------------------------------------------------------------------------
1 | 


--------------------------------------------------------------------------------
/web_application/static/bg.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/web_application/static/bg.jpg


--------------------------------------------------------------------------------
/documentation/project_report.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/documentation/project_report.pdf


--------------------------------------------------------------------------------
/web_application/templates/bg.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/web_application/templates/bg.jpg


--------------------------------------------------------------------------------
/documentation/assets/images/home_page.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/documentation/assets/images/home_page.png


--------------------------------------------------------------------------------
/documentation/assets/images/compare_page.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/documentation/assets/images/compare_page.png


--------------------------------------------------------------------------------
/documentation/assets/images/predict_page.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/documentation/assets/images/predict_page.png


--------------------------------------------------------------------------------
/documentation/assets/images/results_page.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/documentation/assets/images/results_page.png


--------------------------------------------------------------------------------
/web_application/static/comparison_graph.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/web_application/static/comparison_graph.png


--------------------------------------------------------------------------------
/web_application/static/temp/UNADJUSTEDNONRAW_thumb_1d.jpg:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/jenishborah/Tea-Leaf-Disease-Detection/HEAD/web_application/static/temp/UNADJUSTEDNONRAW_thumb_1d.jpg


--------------------------------------------------------------------------------
/data/README.md:
--------------------------------------------------------------------------------
 1 | # Dataset
 2 | 
 3 | The dataset "Identifying Disease in Tea leaves" used for the tea leaf disease detection project can be downloaded from the link below:
 4 | 
 5 | [Download Dataset (tea_leaf_dataset)](https://www.kaggle.com/datasets/shashwatwork/identifying-disease-in-tea-leafs)
 6 | 
 7 | Please make sure to cite the source of the dataset if you use it for any research or projects.
 8 | 
 9 | For more details about the project, refer to the main README.md.
10 | 


--------------------------------------------------------------------------------
/web_application/templates/prediction.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 | <head>
 4 |     <title>Prediction</title>
 5 |     <link rel="stylesheet" type="text/css" href="{{ url_for('static', filename='css/style.css') }}">
 6 | </head>
 7 | <body style="background-image: url('{{ url_for('static', filename='images/nature.jpg') }}');">
 8 |     <h1>Prediction</h1>
 9 |     {% if accuracy and loss %}
10 |         <h3>Accuracy: {{ accuracy }}</h3>
11 |         <h3>Loss: {{ loss }}</h3>
12 |     {% endif %}
13 |     <form method="POST" action="/predict" enctype="multipart/form-data">
14 |         <label for="image">Select Image:</label>
15 |         <input type="file" name="image" id="image" accept="image/*">
16 |         <br><br>
17 |         <input type="submit" value="Predict">
18 |     </form>
19 | </body>
20 | </html>
21 | 


--------------------------------------------------------------------------------
/web_application/templates/comapre.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 | <head>
 4 |     <title>Model Comparison Results</title>
 5 |     <style>
 6 |         body {
 7 |             background-color: #5DADE2;
 8 |         }
 9 | 
10 |         .container {
11 |             width: 800px;
12 |             margin: 0 auto;
13 |             padding: 20px;
14 |             background-color: #F2F3F4;
15 |             border-radius: 5px;
16 |             box-shadow: 0 0 10px rgba(0, 0, 0, 0.2);
17 |         }
18 | 
19 |         h1 {
20 |             text-align: center;
21 |         }
22 | 
23 |         img {
24 |             display: block;
25 |             margin: 0 auto;
26 |             max-width: 600px;
27 |         }
28 |     </style>
29 | </head>
30 | <body>
31 | <div class="container">
32 |     <h1>Model Comparison Results</h1>
33 |     <h2>Comparison Graph:</h2>
34 |     <img src="{{ comparison_graph_path }}" alt="Comparison Graph">
35 | </div>
36 | </body>
37 | </html>
38 | 


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2023 Jenish
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/web_application/static/style.css:
--------------------------------------------------------------------------------
 1 | /* styles for index.html */
 2 | body {
 3 |     font-family: Arial, sans-serif;
 4 |   }
 5 |   
 6 |   .container {
 7 |     max-width: 600px;
 8 |     margin: 0 auto;
 9 |     padding: 20px;
10 |   }
11 |   
12 |   h1 {
13 |     text-align: center;
14 |     margin-bottom: 20px;
15 |   }
16 |   
17 |   form {
18 |     display: flex;
19 |     flex-direction: column;
20 |     align-items: center;
21 |   }
22 |   
23 |   label {
24 |     margin-bottom: 10px;
25 |   }
26 |   
27 |   input[type="text"],
28 |   input[type="number"] {
29 |     padding: 5px;
30 |     margin-bottom: 10px;
31 |     width: 100%;
32 |   }
33 |   
34 |   select {
35 |     padding: 5px;
36 |     margin-bottom: 10px;
37 |     width: 100%;
38 |   }
39 |   
40 |   button {
41 |     padding: 10px 20px;
42 |   }
43 |   
44 |   /* styles for result.html */
45 |   .result-container {
46 |     max-width: 800px;
47 |     margin: 0 auto;
48 |     padding: 20px;
49 |   }
50 |   
51 |   h2 {
52 |     text-align: center;
53 |     margin-bottom: 20px;
54 |   }
55 |   
56 |   img.plot {
57 |     display: block;
58 |     margin: 0 auto;
59 |     width: 100%;
60 |     max-width: 600px;
61 |     border: 1px solid #ddd;
62 |     padding: 5px;
63 |     box-sizing: border-box;
64 |   }
65 |   


--------------------------------------------------------------------------------
/web_application/templates/result.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 | <head>
 4 |     <title>Model Comparison Result</title>
 5 |     <style>
 6 |         body {
 7 |             background-color: #5DADE2;
 8 |         }
 9 | 
10 |         .container {
11 |             width: 500px;
12 |             margin: 0 auto;
13 |             padding: 20px;
14 |             background-color: #F2F3F4;
15 |             background-image: url('background_image.jpg');
16 |             background-size: cover;
17 |             background-position: center;
18 |             border-radius: 5px;
19 |             box-shadow: 0 0 10px rgba(0, 0, 0, 0.2);
20 |         }
21 | 
22 |         h1 {
23 |             text-align: center;
24 |         }
25 | 
26 |         .result {
27 |             text-align: center;
28 |             margin-top: 30px;
29 |         }
30 | 
31 |         .result h2 {
32 |             margin-bottom: 10px;
33 |         }
34 | 
35 |         .result p {
36 |             margin-top: 0;
37 |         }
38 |     </style>
39 | </head>
40 | <body>
41 | <div class="container">
42 |     <h1>Model Comparison Result</h1>
43 |     <div class="result">
44 |         <h2>Accuracy Comparison</h2>
45 |         <p>Model 1 Accuracy: {{ accuracy_model1 }}</p>
46 |         <p>Model 2 Accuracy: {{ accuracy_model2 }}</p>
47 |     </div>
48 |     <div class="result">
49 |         <h2>Loss Comparison</h2>
50 |         <p>Model 1 Loss: {{ loss_model1 }}</p>
51 |         <p>Model 2 Loss: {{ loss_model2 }}</p>
52 |     </div>
53 |     <div class="result">
54 |         <h2>Prediction</h2>
55 |         <p>Prediction: {{ prediction }}</p>
56 |     </div>
57 | </div>
58 | </body>
59 | </html>
60 | 


--------------------------------------------------------------------------------
/web_application/templates/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 | <head>
 4 |     <title>Model Comparison</title>
 5 |     <style>
 6 |         body {
 7 |             background-color: #5DADE2;
 8 |         }
 9 | 
10 |         .container {
11 |             width: 500px;
12 |             margin: 0 auto;
13 |             padding: 20px;
14 |             background-color: #F2F3F4;
15 |             border-radius: 5px;
16 |             box-shadow: 0 0 10px rgba(0, 0, 0, 0.2);
17 |         }
18 | 
19 |         h1 {
20 |             text-align: center;
21 |         }
22 | 
23 |         form {
24 |             text-align: center;
25 |         }
26 | 
27 |         .form-group {
28 |             margin-bottom: 20px;
29 |         }
30 | 
31 |         .form-group label {
32 |             display: block;
33 |             font-weight: bold;
34 |         }
35 | 
36 |         .form-group input,
37 |         .form-group select {
38 |             width: 100%;
39 |             padding: 5px;
40 |             border-radius: 3px;
41 |             border: 1px solid #ccc;
42 |         }
43 | 
44 |         .form-group input[type="submit"] {
45 |             background-color: #3498DB;
46 |             color: #fff;
47 |             padding: 10px 20px;
48 |             border: none;
49 |             border-radius: 3px;
50 |             cursor: pointer;
51 |         }
52 |     </style>
53 | 
54 |     <title>Model Selection</title>
55 | 
56 | </head>
57 | <body style="background-image: url('{{ url_for('static', filename='bg.jpg') }}');">
58 |     <h1>Model Selection</h1>
59 |     <form method="POST" action="/predict">
60 |         <label for="model">Select Model:</label>
61 |         <select name="model" id="model">
62 |             <option value="cnn">Custom Proposed CNN Model</option>
63 |             <option value="inception">Customized Inception V3</option>
64 |             <option value="vgg16">Customized VGG16</option>
65 |             <option value="nasnet">Customized NasNetMobile</option>
66 |         </select>
67 | 
68 |         <br><br>
69 | 
70 |         <label for="batch_size">Select Batch Size:</label>
71 |         <select name="batch_size" id="batch_size">
72 |             <option value="16">16</option>
73 |             <option value="32">32</option>
74 |             <option value="128">128</option>
75 |         </select>
76 | 
77 |         <br><br>
78 | 
79 |         <label for="epochs">Select Number of Epochs:</label>
80 |         <input type="number" name="epochs" id="epochs" min="1" max="100" value="10">
81 | 
82 |         <br><br>
83 | 
84 |         <input type="submit" value="Train and Predict">
85 |     </form>
86 | </body>
87 | </html>
88 | 
89 | 
90 | 


--------------------------------------------------------------------------------
/requirements.txt:
--------------------------------------------------------------------------------
  1 | absl-py==1.4.0
  2 | aiohttp==3.8.4
  3 | aiohttp-retry==2.8.3
  4 | aiosignal==1.3.1
  5 | appdirs==1.4.4
  6 | astunparse==1.6.3
  7 | async-timeout==4.0.2
  8 | attrs==23.1.0
  9 | beautifulsoup4==4.12.2
 10 | cachetools==5.3.0
 11 | certifi==2022.12.7
 12 | cffi==1.15.1
 13 | charset-normalizer==3.1.0
 14 | click==8.1.3
 15 | colorama==0.4.6
 16 | comtypes==1.1.14
 17 | contourpy==1.0.7
 18 | cycler==0.11.0
 19 | fastjsonschema==2.17.1
 20 | filelock==3.12.0
 21 | Flask==2.2.3
 22 | flatbuffers==23.3.3
 23 | fonttools==4.39.3
 24 | frozenlist==1.3.3
 25 | gast==0.4.0
 26 | gevent==22.10.2
 27 | google-auth==2.17.1
 28 | google-auth-oauthlib==1.0.0
 29 | google-pasta==0.2.0
 30 | greenlet==2.0.2
 31 | grpcio==1.53.0
 32 | h5py==3.8.0
 33 | hmmlearn==0.3.0
 34 | idna==3.4
 35 | itsdangerous==2.1.2
 36 | jax==0.4.8
 37 | Jinja2==3.1.2
 38 | joblib==1.2.0
 39 | keras==2.12.0
 40 | kiwisolver==1.4.4
 41 | libclang==16.0.0
 42 | Markdown==3.4.3
 43 | MarkupSafe==2.1.2
 44 | matplotlib==3.7.1
 45 | ml-dtypes==0.0.4
 46 | MouseInfo==0.1.3
 47 | mpmath==1.3.0
 48 | multidict==6.0.4
 49 | networkx==3.1
 50 | numpy==1.23.5
 51 | oauthlib==3.2.2
 52 | opencv-python==4.7.0.72
 53 | opt-einsum==3.3.0
 54 | packaging==21.3
 55 | pandas==2.0.1
 56 | Pillow==9.5.0
 57 | protobuf==4.22.1
 58 | psutil==5.9.5
 59 | pyasn1==0.4.8
 60 | pyasn1-modules==0.2.8
 61 | PyAudio==0.2.13
 62 | PyAutoGUI==0.9.53
 63 | pycparser==2.21
 64 | pydot==1.4.2
 65 | pyee==8.2.2
 66 | PyGetWindow==0.0.9
 67 | Pygments==2.15.1
 68 | pyjokes==0.6.0
 69 | PyJWT==2.7.0
 70 | PyMsgBox==1.0.9
 71 | pyparsing==3.0.9
 72 | pyperclip==1.8.2
 73 | pypiwin32==223
 74 | PyRect==0.2.0
 75 | pyrsistent==0.19.3
 76 | PyScreeze==0.1.28
 77 | python-dateutil==2.8.2
 78 | python-http-client==3.3.7
 79 | pyttsx3==2.90
 80 | pytweening==1.0.4
 81 | pytz==2023.3
 82 | pywhatkit==5.4
 83 | pywin32==306
 84 | PyYAML==6.0
 85 | pyzmq==25.1.0
 86 | requests==2.28.2
 87 | requests-oauthlib==1.3.1
 88 | rsa==4.9
 89 | scikeras==0.4.1
 90 | scikit-learn==1.2.2
 91 | scipy==1.10.1
 92 | seaborn==0.12.2
 93 | sendgrid==6.10.0
 94 | sentry-sdk==1.22.2
 95 | six==1.16.0
 96 | sklearn==0.0.post1
 97 | soundfile==0.12.1
 98 | soupsieve==2.4
 99 | SpeechRecognition==3.10.0
100 | starkbank-ecdsa==2.2.0
101 | sympy==1.12
102 | tensorboard==2.12.1
103 | tensorboard-data-server==0.7.0
104 | tensorboard-plugin-wit==1.8.1
105 | tensorflow==2.12.0
106 | tensorflow-estimator==2.12.0
107 | tensorflow-intel==2.12.0
108 | tensorflow-io-gcs-filesystem==0.31.0
109 | termcolor==2.2.0
110 | thop==0.1.1.post2209072238
111 | threadpoolctl==3.1.0
112 | tinycss2==1.2.1
113 | torch==2.0.1
114 | torchvision==0.15.2
115 | tornado==6.3.2
116 | tqdm==4.65.0
117 | traitlets==5.9.0
118 | twilio==8.4.0
119 | typing_extensions==4.5.0
120 | tzdata==2023.3
121 | ultralytics==8.0.99
122 | urllib3==1.26.15
123 | webencodings==0.5.1
124 | websockets==10.4
125 | Werkzeug==2.2.3
126 | wikipedia==1.4.0
127 | wrapt==1.14.1
128 | yarl==1.9.2
129 | zipp==3.15.0
130 | zope.event==4.6
131 | zope.interface==6.0
132 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # Tea-Leaf-Disease-Detection
 2 | This repo contains the code and documentation for my final year project on tea leaf diseases using deep learning as a partial fulfillment of my degree of master in science in information technology from the department of computer science, Gauhati university, Ghy-14
 3 | 
 4 | # !Important!
 5 | The complete folders including `saved_models` are in master branch instead of main branch
 6 | 
 7 | 
 8 | ## Dataset
 9 | I used a publicly available dataset for tea leaf disease detection. The dataset is available in the `data` directory.
10 | 
11 | ## Notebooks
12 | 
13 | We trained four different machine learning models to classify tea leaf diseases. The models are saved in the `saved_models` directory. For more details about the models and their performance, refer to the project proposal/project report in the `documentation` directory.
14 | 
15 | 
16 | - `VGG16withCustomLayers.ipynb`: Jupyter Notebook for tea leaf disease detection using transfer learing technique: base model as VGG16.
17 | - `NasNetWithCustomLayers.ipynb`: Jupyter Notebook for tea leaf disease detection using transfer learing technique: base model as NasNetMobile.
18 | - `InceptionV3.ipynb`: Jupyter Notebook for tea leaf disease detection using transfer learing technique: base model as InceptionV3.
19 | - `Mysequential.ipynb`: Jupyter Notebook for tea leaf disease detection using sequentail personal customized CNN model.
20 | 
21 | 
22 | 
23 | ## Saved Models
24 | The trained models are saved in the `saved_models` directory.
25 | 
26 | ## Web Application
27 | We have developed a web application to allow users to interact with the trained models and get predictions for tea leaf diseases.
28 | 
29 | 
30 | ### Web Application Features
31 | 
32 | - Home Page: The home page allow user to upload a dataset path or link and allow user to choose two trained models among the four machine learning models.
33 | 
34 | - Comapre Page: The compare page shows the comparesion graph of the two selected models by the user in the home page along with the hyparperameters. 
35 | 
36 | - Predict  Page: Users can select any model among the four train/saved model and give input images of tea leaf to detect the tea leaf disease
37 | 
38 | - Result Page: Users get predictions for the type of disease affecting the tea leaves.
39 | 
40 | ### Screenshots
41 | 
42 | ![Home Page](documentation/assets/images/home_page.png)
43 | 
44 | ![Compare Page](documentation/assets/images/compare_page.png)
45 | 
46 | ![Predict Page](documentation/assets/images/predict_page.png)
47 | 
48 | ![Result Page](documentation/assets/images/results_page.png)
49 | 
50 | 
51 | ## Documentation
52 | The `documentation` directory contains my project proposal, Report and assests. For more detailed information about the project, you can refer to the following documents:
53 | 
54 | - [Project Proposal](documentation/project_proposal.pdf)
55 | - [Project Report](documentation/project_report.pdf)
56 | - [Assets](documentation/assets)
57 | 
58 | ## How to Use
59 | 
60 | 
61 | ### Installation
62 | 
63 | Make sure you have the required version of Python (e.g., Python 3.7 or later) installed before proceeding.
64 | 
65 | 1. To use this project, you need to have Python installed on your system.
66 |    
67 | 2. Install the required dependencies using the following command: `pip install -r requirements.txt`
68 | 
69 | 3. Open your terminal, navigate to where you want to save the project and and Clone this repository using `git clone https://github.com/jenishborah/Tea-Leaf-Disease-Detection`
70 | 
71 | 4. Open and run the Jupyter Notebooks in the `notebooks` directory in the specified order.
72 | 
73 | 
74 | 
75 | 
76 | 
77 | 
78 | #### Running the Web Application
79 | 
80 | 1. Navigate to the `web_app` directory.
81 | 
82 | 2. Run the Flask application: `main.py`
83 |    
84 | 3. Open your web browser and go to `http://localhost:5000` to access the web application.
85 | 
86 | 
87 | 
88 | ## License
89 | This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
90 | 
91 | Copyright (c) 2023 <Jenish> Jenish A borah  </Jenish>
92 | 


--------------------------------------------------------------------------------
/web_application/main.py:
--------------------------------------------------------------------------------
  1 | from flask import Flask, render_template, request
  2 | import os
  3 | import matplotlib.pyplot as plt
  4 | import numpy as np
  5 | import cv2
  6 | import tensorflow as tf
  7 | from tensorflow.keras.preprocessing.image import ImageDataGenerator
  8 | from tensorflow.keras.applications import InceptionV3, VGG16, ResNet50, MobileNetV2
  9 | from tensorflow.keras.layers import GlobalAveragePooling2D, Dense
 10 | from tensorflow.keras.models import Model
 11 | 
 12 | app = Flask(__name__)
 13 | 
 14 | # Define the load_dataset function before the compare function
 15 | def load_dataset(dataset_path):
 16 |     if not os.path.exists(dataset_path):
 17 |         raise FileNotFoundError("Dataset directory does not exist: " + dataset_path)
 18 | 
 19 |     # Load the dataset from the directory
 20 |     return dataset_path
 21 | 
 22 | 
 23 | @app.route('/')
 24 | def index():
 25 |     return render_template('index.html')
 26 | 
 27 | 
 28 | @app.route('/compare', methods=['POST'])
 29 | def compare():
 30 |     # Get the selected model names
 31 |     model1 = request.form.get('model1')
 32 |     model2 = request.form.get('model2')
 33 | 
 34 |     # Get the selected parameters
 35 |     epochs = int(request.form.get('epochs'))
 36 |     batch_size = int(request.form.get('batch_size', 16))
 37 | 
 38 | 
 39 |     # Get the form data
 40 |     dataset_link = request.form.get('dataset_link')
 41 |     dataset_path = request.form.get('dataset_path')
 42 | 
 43 |     if dataset_path is None:
 44 |         return "Please provide a valid dataset path."
 45 | 
 46 |     # Load the dataset
 47 |     dataset = load_dataset(dataset_path)
 48 | 
 49 |     # Preprocess the dataset based on the selected models
 50 |     preprocessed_data1, preprocessed_data2 = preprocess_data(dataset, model1, model2)
 51 | 
 52 |     # Train and compare the selected models
 53 |     if model1 != model2:
 54 |         history_model1, history_model2 = train_and_compare_models(model1, model2, epochs, batch_size, preprocessed_data1, preprocessed_data2)
 55 |         generate_comparison_graph(history_model1, history_model2)
 56 |     else:
 57 |         return "Please select different models for comparison."
 58 | 
 59 |     # Return the path to the generated comparison graph
 60 |     comparison_graph_path = os.path.join('static', 'comparison_graph.png')
 61 |     return render_template('result.html', comparison_graph_path=comparison_graph_path)
 62 | 
 63 | 
 64 | def preprocess_data(dataset, model1, model2):
 65 |     preprocessed_data1 = []
 66 |     preprocessed_data2 = []
 67 | 
 68 |     if model1 == 'custom-nasnetmobile':
 69 |         # Implement the preprocessing steps for Custom NASNetMobile
 70 |         preprocessed_data1 = dataset
 71 |     elif model1 == 'custom-vgg16':
 72 |         # Implement the preprocessing steps for Custom VGG16
 73 |         preprocessed_data1 = dataset
 74 |     elif model1 == 'custom-inceptionv3':
 75 |         # Implement the preprocessing steps for Custom InceptionV3
 76 |         preprocessed_data1 = dataset
 77 | 
 78 |     if model2 == 'custom-nasnetmobile':
 79 |         # Implement the preprocessing steps for Custom NASNetMobile
 80 |         preprocessed_data2 = dataset
 81 |     elif model2 == 'custom-vgg16':
 82 |         # Implement the preprocessing steps for Custom VGG16
 83 |         preprocessed_data2 = dataset
 84 |     elif model2 == 'custom-inceptionv3':
 85 |         # Implement the preprocessing steps for Custom InceptionV3
 86 |         preprocessed_data2 = dataset
 87 | 
 88 |     return preprocessed_data1, preprocessed_data2
 89 | 
 90 | 
 91 | def load_and_preprocess_dataset(dataset_path, batch_size, validation_split):
 92 |     # Create an ImageDataGenerator for data preprocessing
 93 |     datagen = ImageDataGenerator(
 94 |         rescale=1./255,
 95 |         validation_split=validation_split
 96 |     )
 97 | 
 98 |     # Load and preprocess the training dataset
 99 |     train_data = datagen.flow_from_directory(
100 |         dataset_path,
101 |         target_size=(224, 224),
102 |         batch_size=batch_size,
103 |         class_mode='categorical',
104 |         subset='training'
105 |     )
106 | 
107 |     # Load and preprocess the validation dataset
108 |     val_data = datagen.flow_from_directory(
109 |         dataset_path,
110 |         target_size=(224, 224),
111 |         batch_size=batch_size,
112 |         class_mode='categorical',
113 |         subset='validation'
114 |     )
115 | 
116 |     return train_data, val_data
117 | 
118 | 
119 | def build_model(model_name):
120 |     base_model = None
121 | 
122 |     if model_name == 'custom-nasnetmobile':
123 |         base_model = CustomNASNetMobile(include_top=False, weights='imagenet', input_shape=(224, 224, 3))
124 |     elif model_name == 'custom-vgg16':
125 |         base_model = CustomVGG16(include_top=False, weights='imagenet', input_shape=(224, 224, 3))
126 |     elif model_name == 'custom-inceptionv3':
127 |         base_model = CustomInceptionV3(include_top=False, weights='imagenet', input_shape=(224, 224, 3))
128 | 
129 |     # Add custom classification layers on top of the base model
130 |     x = base_model.output
131 |     x = GlobalAveragePooling2D()(x)
132 |     x = Dense(1024, activation='relu')(x)
133 |     predictions = Dense(8, activation='softmax')(x)
134 | 
135 |     # Create the final model
136 |     model = Model(inputs=base_model.input, outputs=predictions)
137 | 
138 |     # Compile the model
139 |     model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])
140 | 
141 |     return model
142 | 
143 | 
144 | def train_and_compare_models(model1, model2, epochs, batch_size, data1, data2):
145 |     # Load and preprocess the datasets
146 |     train_data1, val_data1 = load_and_preprocess_dataset(data1, batch_size, 0.2)
147 |     train_data2, val_data2 = load_and_preprocess_dataset(data2, batch_size, 0.2)
148 | 
149 |     # Build the models
150 |     model1 = build_model(model1)
151 |     model2 = build_model(model2)
152 | 
153 |     # Train the models
154 |     history_model1 = model1.fit(train_data1, epochs=epochs, validation_data=val_data1)
155 |     history_model2 = model2.fit(train_data2, epochs=epochs, validation_data=val_data2)
156 | 
157 |     return history_model1, history_model2
158 | 
159 | 
160 | def generate_comparison_graph(history_model1, history_model2):
161 |     # Get the training and validation accuracy
162 |     train_acc_model1 = history_model1.history['accuracy']
163 |     val_acc_model1 = history_model1.history['val_accuracy']
164 |     train_acc_model2 = history_model2.history['accuracy']
165 |     val_acc_model2 = history_model2.history['val_accuracy']
166 | 
167 |     # Get the training and validation loss
168 |     train_loss_model1 = history_model1.history['loss']
169 |     val_loss_model1 = history_model1.history['val_loss']
170 |     train_loss_model2 = history_model2.history['loss']
171 |     val_loss_model2 = history_model2.history['val_loss']
172 | 
173 |     # Plot the training and validation accuracy
174 |     plt.figure(figsize=(12, 6))
175 |     plt.subplot(1, 2, 1)
176 |     plt.plot(train_acc_model1, label='Model 1 Training Accuracy')
177 |     plt.plot(val_acc_model1, label='Model 1 Validation Accuracy')
178 |     plt.plot(train_acc_model2, label='Model 2 Training Accuracy')
179 |     plt.plot(val_acc_model2, label='Model 2 Validation Accuracy')
180 |     plt.xlabel('Epochs')
181 |     plt.ylabel('Accuracy')
182 |     plt.title('Training and Validation Accuracy')
183 |     plt.legend()
184 | 
185 |     # Plot the training and validation loss
186 |     plt.subplot(1, 2, 2)
187 |     plt.plot(train_loss_model1, label='Model 1 Training Loss')
188 |     plt.plot(val_loss_model1, label='Model 1 Validation Loss')
189 |     plt.plot(train_loss_model2, label='Model 2 Training Loss')
190 |     plt.plot(val_loss_model2, label='Model 2 Validation Loss')
191 |     plt.xlabel('Epochs')
192 |     plt.ylabel('Loss')
193 |     plt.title('Training and Validation Loss')
194 |     plt.legend()
195 | 
196 |     # Save the comparison graph
197 |     comparison_graph_path = os.path.join('static', 'comparison_graph.png')
198 |     plt.savefig(comparison_graph_path)
199 |     plt.close()
200 | 
201 |     return comparison_graph_path
202 | 
203 | 
204 | if __name__ == '__main__':
205 |     app.run(debug=True)
206 | 


--------------------------------------------------------------------------------