ARTICLE AD BOX
I am processing a video and need to extract frames that represent
consecutive portions of the scene.
The requirement is that each extracted frame should correspond to
the next part of the scene relative to the previous frame, so that
the frames form a continuous sequence without overlapping visual
content.
Example conceptually:
Frame 1 → first portion of the scene
Frame 2 → next portion immediately after frame 1
Frame 3 → next portion after frame 2
Currently I am extracting frames at fixed time intervals, but this
often produces frames with overlapping content because the camera
movement in the video is not constant.
What would be a recommended approach to extract frames based on
visual change or scene progression rather than fixed timestamps?
Is there a way to do this using AVFoundation ?
