Programmatically getting still frames from cameras for post-processing?

Question

Python

There is a video file . It is
necessary to cut it into a freeze frame with a period of 0.5 seconds and save them for further processing. Later,
this will all develop as a project for face recognition
in Google, only ready-made applications are offered, but I need a software implementation.
Preferably in python, but sharp is also suitable
Advise literature, found little specific on this subject
. Ideally, I understand that you need to make a real-time system, but this is enough to start.

Answer 1

If you only need to decompose into frames, then ffmpeg will do it quickly and for free.
There is no point in doing parsing inside python, do a system call

Answer 2

Check out the sample here https://stackoverflow.com/questions/33311153/pytho...

Answer 3

DirectShow interfaces provide access to each frame (usually 30 frames per second for cameras).
https://docs.microsoft.com/en-us/windows/desktop/d...