🩹 Wound Detection AI

A deep learning-powered image classification app that identifies wound types from medical images. Built using PyTorch and Streamlit, this project allows users to upload images and receive predictions across 10 common wound categories.

---

🔍 Features

• ✅ Classifies wounds into 10 types:
  • Abrasions, Bruises, Burns, Cuts, Diabetic Wounds, Lacerations, Normal, Pressure Wounds, Surgical Wounds, Venous Wounds
• 🧠 Uses ResNet50 + transfer learning
• ⚡ Fast and lightweight with Streamlit
• 📦 Easy to retrain or integrate into other workflows

---

🧪 Tech Stack

• Python
• PyTorch
• Torchvision
• Streamlit
• PIL (Python Imaging Library)

---

🚀 Getting Started

1. Clone the repository:

   git clone https://github.com/kalunkuo/wound-detection-ai.git
   cd wound-detection-ai
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Place your dataset in the dataset/ folder (each wound type should be a folder with images inside).

• Kaggle: https://www.kaggle.com/datasets/ibrahimfateen/wound-classification/data

4. Train the model:

   python train.py
   

5. Run the Streamlit app:

   streamlit run wound_app.py