Primitive set of tools for video content analysis
Requirements software side of things You'll need ffmpeg, this code and some instance empovered with Ubuntu 16.04-18.04 (tested) sources We start with teletext capable streams that should be used as input as [STREAM URL], voice recognition is planned to add later.
Usage content preparing Let's start from raw data preparing Navigate to the desired location in your server (cd /smth/) and execute in the command line the following:
ffmpeg -loglevel quiet -txt_format text -txt_page 888 -i [STREAM URL] -f srt -
GOP structure check Sometimes you may need to analyze how group of pictures look like in particular file or stream. That python script will assist you here.