Creating depth maps from stereo video?

I'm doing some experimenting.

So I set up two identical cameras side-by-side, recorded some test footage. (And also have synchronised the footage based on the audio)

At this point I would like to create a depth map, for one or other (or both?) of the two video streams.

I found some high-end plug-ins that would do this (too expensive )

And also managed to find some research papers detailing some of the algorithms that can be used, so I could, in theory, create my own tool to do this. But that would require a certain amount of time and effort.


So the question is, does anyone know of a middle ground? Has anyone played around with this sort of thing before, and have any tips, or know of any good software tools?


  • Triem23Triem23 Moderator

    Oh, boy....

    Mocha Pro has tools for Stereo... I guess you'd have to roto varied things?

    Generating a depth map from stereo footage is out of my experience. Oddly enough, I'm about to work on an outline for some ideas to do stereo animation setups in Hitfilm, but it doesn't cover your question...

    Good question, though. I'll be interested to see if anyone else has anything to add.


  • My first thought would be OpenCV and Python to roll your own. OpenCV should take alot of the grunt work out of it.

    Doing a quick search found this with a lot of command line utilities for doing different things

  • @Aladdin4d I started playing around with OpenCV a bit, seems to do the trick. I will definitely check out the stuff on that blog though. At the very least, it will give me a reference so I can compare output from my home-made python scripts to something else


  • If you do manage to make it work can you share it with us? I am very interested as to how this would work.

Sign in to comment

Leave a Comment