Damage Detection with YOLOv8

A Streamlit web application for detecting damage in images using a pre-trained YOLOv8 model. Users can upload an image, view the detected damage with labeled bounding boxes (showing only damage types), and see detailed detection data in a table.

Features

• Upload images (JPG, JPEG, PNG) for damage detection.
• Display annotated images with bounding boxes labeled by damaged.
• Show a table with detection details: Detection #, Damaged, Confidence, and Bounding Box coordinates.
• Built with Streamlit for an interactive web interface.
• Uses a custom YOLOv8 model (best.pt) for damage detection.

Prerequisites

• Python 3.8 or higher
• A trained YOLOv8 model file (best.pt) in the project directory
• Git (for cloning the repository)

Installation

1. Clone the repository:

   git clone https://github.com/omkarjavali/container_damage_detection.git
   cd container_damage_detection

2. Create a virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Ensure the YOLOv8 model:

   • Place your best.pt file in the project root or update the path in app.py if stored elsewhere.

Usage

Run the Streamlit app:

streamlit run app.py

Access the app:

• Open your browser and go to http://localhost:8501
• Upload an image to detect damage
• View the annotated image with damage labels and a table of detection details below

Project Structure

your-repo-name/
│
├── home.py              # Main Streamlit application
├── pages
    ├── result.py
├── best.pt              # YOLOv8 model file (not tracked in Git)
├── requirements.txt     # Python dependencies
├── README.md            # Project documentation

Notes

• The app assumes best.pt outputs bounding boxes with class labels. Adjust app.py if your model has a different format.
• For large images, consider adding resizing logic in app.py to improve performance.
• To customize the UI (e.g., bounding box colors, table format), modify app.py.

Contributors

• https://github.com/iamsonuram
• https://github.com/D-6503

Acknowledgments

• Built with Streamlit and Ultralytics YOLOv8.
• Thanks to the open-source community for robust tools!