Wood Surface Anomaly Detection and Segmentation

This project aims to detect and segment anomalies on wood surfaces using unsupervised deep learning models. It includes a FastAPI-based backend and a React-based frontend. The following instructions will help you set up and run the project locally.

---

📁 Clone the Repository

git clone https://github.com/melisguclu/anomaly-detection.git
cd wood-anomaly-detection

---

🧪 Backend Setup (FastAPI)

Step 1: Download Pretrained Model

Download the pretrained PaDiM model manually from the following Google Drive link:

🔗 Download PaDiM model

After downloading, unzip the file and place it into the following directory inside the project:

models/padim/train_wood.pkl

Do not unzip it yet. Further steps will handle that automatically if necessary.

Step 2: Set Up Python Environment

Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # For Unix/MacOS
venv\Scripts\activate    # For Windows

Step 3: Install Dependencies

pip install --upgrade pip
pip install -r backend/requirements.txt

Step 4: Run the Backend Server

Navigate into the backend folder and start the FastAPI development server:

cd backend
uvicorn main:app --reload

By default, the API will be available at:

http://127.0.0.1:8000

You can view the automatically generated API docs at:

http://127.0.0.1:8000/docs

---

🚀 Frontend Setup (React)

Step 1: Install Node Modules

Navigate to the frontend directory and install the dependencies:

cd ../frontend
npm install

Step 2: Run the Development Server

npm run dev

This will start the React development server, typically available at:

http://localhost:5173

Make sure the backend is running simultaneously for the application to function properly.

---

🔗 Dataset Structure

Ensure the dataset is placed under the following structure:

wood_dataset/
└── wood/
    ├── train/
    │   └── good/
    ├── test/
    │   ├── good/
    │   └── defect/
    └── ground_truth/
        └── defect/

This dataset structure is crucial for training and evaluating the models.

---

🌟 Notes

• Ensure all Python and Node dependencies are correctly installed.
• The pretrained model must be manually downloaded and placed into the correct folder.
• Backend and frontend must run concurrently for proper functionality.

---

✉️ Contact

For any issues or questions, please open an issue on the GitHub repository or contact the contributors.