I would like to extract pixels associated with specific macroblocks in a video frame.
I can extract macroblock types and identity if one is 8x16 rather than 16x16, for example, but how do I map from this information to the frame pixels themselves? When I extract the macroblock type information (using -debug mb_type), I'm also confused as to why I don't see 8x16 macroblock types adjacent to one another. I thought frames were first divided into 16x16 blocks, then subdivided into, for example 8x16. Is this incorrect? Thanks for any help offered! _______________________________________________ ffmpeg-user mailing list [email protected] https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
