Video Parser

The Video Parser is a Python script that converts a video file into a time series data file, extracting the time and value information from the video frames.

Prerequisites

• Python 3.7 or higher
• Docker (for Docker usage)

Installation

Docker Installation

1. Clone the repository and navigate to the project directory (this way you have the test .mp4 file):

git clone https://github.com/stuart-farris/video_parser.git
cd video_parser

2. Install Docker on your system following the official Docker installation instructions: Docker Installation Guide

Non-Docker Installation

1. Install the python package:

python3 -m pip install video_parser 'git+https://github.com/stuart-farris/video_parser.git@8a4ca186168f7681c33e12b8520e4d2c5d5b1f71'

2. Clone the repository and navigate to the project directory:

git clone https://github.com/stuart-farris/video_parser.git
cd video_parser

Usage

Docker Usage

1. Run the video_parser using the following Docker command:

docker run \
  -v $PWD/Problem_1/Test_Video.mp4:/app/video.mp4 \
  -v $PWD/results:/app/results \
  sfarris1994/video_parser video_parser

You can also replace $PWD/Problem_1/Test_Video.mp4 with the path to your video file. The video file should be mounted as a volume inside the Docker container.

The time series data will be extracted from the video file, and a file named output.csv will be created in the current directory.

To run the video parser with GPU acceleration using Docker, use the --gpu argument, as follows:

docker run \
  -v $PWD/Problem_1/Test_Video.mp4:/app/video.mp4 \
  -v $PWD/results:/app/results \
  --gpus 1 \
  sfarris1994/video_parser video_parser --gpu

Non-Docker Usage

1. Navigate to the project directory:

cd video_parser

2. Run the video_parser script with the following command:

video_parser --video Problem_1/Test_Video.mp4

You can also replace Problem_1/Test_Video.mp4 with the path to your video file.

The time series data will be extracted from the video file, and a file named output.csv will be created in the current directory.

To run the video parser with GPU acceleration, use the --gpu argument, as follows:

python video_parser.py Problem_1/Test_Video.mp4 --gpu --video Problem_1/Test_Video.mp4

License

This project is licensed under the MIT License - see the LICENSE file for details.