I'd like to use python to take a video (eventually a stream) and map each frame to an RGB array of pixels.

My first thought was to use ffmpeg bindings, and in memory (if possible) convert it to png to read each pixel. However, I think that this is redundant to the purpose of the decoder.

It seems like this may be the most basic entry into video processing, if someone could point me to the library or approach that would accomplish this I would be grateful.


1 Answer 1


You can do it with OpenCV library. Here is OpenCV link for Python library.

  • $\begingroup$ Thank you. There are so many tools it can be a bit of an information overload when searching for something as a general starting point. OpenCV was one that came up in my searching but I haven't gotten into investigating it yet. $\endgroup$
    – Justin
    Sep 11, 2018 at 22:34
  • $\begingroup$ If your plan is to work with image processing, video processing, matrix you should dive into OpenCV. Feel free to ask when you stack. $\endgroup$
    – ITm
    Sep 12, 2018 at 8:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.