🩺 Diabetic Retinopathy Detection

A deep learning project to detect the severity of diabetic retinopathy using retinal fundus images. The model classifies images into 5 stages: No DR, Mild, Moderate, Severe, and Proliferative DR, using transfer learning with ResNet18.

---

📁 Project Structure

colored_images/
├── Mild/
├── Moderate/
├── No_DR/
├── Proliferate_DR/
└── Severe/

templates/
└── index.html

train.csv
train.ipynb
main.py
retino_model.h5 (Generated after training)

---

🔍 Dataset

link-https://www.kaggle.com/code/kushalkumar8906kumar/hiee-project/notebook

• Directory: colored_images/ with subfolders for each DR category
• Labels: Provided in train.csv
• Classes:
  • No_DR
  • Mild
  • Moderate
  • Severe
  • Proliferate_DR

---

🧠 Model Details

• Framework: TensorFlow / Keras
• Architecture: ResNet18 via transfer learning
• Classification Type: Multiclass (5 classes)
• Final Model Output: retino_model.h5
• Achieved Accuracy: 69%

---

🚀 How to Run

Step 1: Train the Model

jupyter notebook train.ipynb

Step 2: Start the Flask Web App

python main.py

Then go to http://127.0.0.1:5000/ in your browser.

✅ Features Classifies 5 stages of diabetic retinopathy

Transfer learning with ResNet18

Web-based prediction interface using Flask

Real-world medical dataset with labeled fundus images