The codebase for a tool that automatically grabs the intro scenes from a tv series and creates a single video.
Requires FFMpeg/Avconv and required libraries. Maven.
- Generate fingerprints for audio from 3 episodes from a series.
- Determine which noise is the theme tune.
- Fingerprint all episodes listening for the theme tune.
- Slice the video files from start to time where match was found for fingerprint.
- Joing all output files together.
Steps 1 & 2 are currently not implemented and instead a wav of the themetune is used. Feel free to PR it.
Here is some output of the application https://www.youtube.com/watch?v=wCdE_NCNMrI&t=322s
Also I just wrote this all out in a couple of evenings whilst watching The Office haha... It totally works well but the code is garbage. I will refactor it at some point. I'll probably launch it as service in AWS when I'm finished.