Parkour Automations

Tech used: Python, Scikit-Image, OpenCV, ffmpeg

Script to identify cut points in Minecraft parkour footage and automatically cut between fixed points based on the information in different clips. I designed the algorithm used to cut videos and tuned parameters for machine vision libraries to optimize it for real footage recorded from the game.

It takes uncut footage with long delays in it like this:

And cuts it into smooth footage like this, intelligently removing sections that aren't in use: