Video Beer Segmenter

Overview

Project for the course of Signal Image Video regarding a simple segmenter to detect beer foam through image processing techniques. This task was previously solved using AI.

Notebook

The project includes a Jupyter Notebook that provides an in-depth explanation of the methodology used for video processing. It demonstrates:

• Preprocessing: How video frames are extracted and converted for processing.
• Segmentation Techniques: Methods applied to identify beer in the video frames.
• Differences with AI solution: discussion on the differences of the two approaches

To run the notebook:

jupyter notebook beer_segmenter.ipynb

Installation

1. Clone the Repository

git clone https://github.com/your-repo/video-beer-segmenter.git
cd video-beer-segmenter

2. Create a Virtual Environment

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

3. Install Dependencies

pip install -r requirements.txt

Usage

Running the Video Beer Segmenter

Modify the video_beer_segmenter.py script to specify the input video path and the output video name inside the main function:

if __name__ == "__main__":
    input_video = "path/to/input_video.mp4"  # Modify this
    output_video = "output_video.mp4"  # Modify this
    process_video(input_video, output_video)

Then run:

python video_beer_segmenter.py

Dependencies

This project requires the following packages:

• numpy
• opencv-python
• matplotlib
• Pillow
• tqdm

License

MIT License

Author

Alessandro De Vidi - aledevv