Revision history [back]

It is actually simple, read in the frames, one by one, and than save them into a compressed format. For video this can be motion JPEG, for imagery it can be JPEG.

Just look at the video and image reading and writing interfaces at the docs : docs.opencv.org