Sensor Image Scratch Detection Web Interface

This repository provides a web interface for detecting scratches in sensor images using a UNET-based model. This project is currently hosted by CERN, and some files are confidential due to project sensitivity.

Author

Abhinav Raj Gupta

Installation

To install the required dependencies, use the following command:

 pip install -r requirements.txt

Setup and Usage

Follow these steps to set up and run the web interface:

1. Clone the repository or download the ZIP file:

  git clone <repository-url>

2. Navigate to the project directory:

  cd <project-directory>

3. Run the application:

  flask run

4. Access the web interface:

Open your local web browser and navigate to http://127.0.0.1:5000.

5. Upload and Analyze Images:

• Use the Upload button to select and upload images.
• Uploaded images are saved in instance/upload/, organized by date and time.
• The UNET model processes the images, and annotated results are stored in static/, also organized by date and time.

6. View and Download Results:

• Annotated images are displayed on the web interface.
• Use the Download Label Files button to download the labels as a ZIP file.

7. Filter Options:

• A dropdown menu is available to filter images as needed.

Note: This repository does not include certain project files due to their confidentiality, as this is part of an official CERN project.