Thanks for chiming in, inhahe. On Tue, 6 Aug 2024 at 16:20, inhahe <inh...@gmail.com> wrote: > > On Tue, Aug 6, 2024 at 8:41 AM Rob Hallam <ffm...@roberthallam.com> wrote: > > I'd like to programmatically detect the 'busiest' parts of a video- ie > > the most visually active areas. I am leaving audio aside for the > > purposes of considering this. > > > > I figured it might be possible by looking at one / more of: > > > > - the bitrate for VBR videos -- a higher bitrate for a given segment > > will tend to be associated with more activity (other things being > > equal) > > - frame differences -- count pixels/blocks which differ, average over > > a time segment > > - optical flow? (higher flow values = more activity) I'm not too > > familiar with this
> The second thing I thought of was your second suggestion, > but ... [snip] ... I wouldn't try this way. Agreed. > I just looked up "optical flow," and it seems to be more or less a > description of why the second approach wouldn't work, Yeah that's fair enough. I am hoping there might be some relatively quick way of working out changes, motion, flow (etc), since encoders would have to do this anyway. I mean, it is possible to re-encode a CBR video as VBR+constant perceptual quality and then do the same bitrate segmentation, but that approach seems very inefficient! Part of the trouble is I know only a little about the techniques of working with videos in this way, so I'm not sure of the unknown unknowns / vulnerable to the anosognosic's dilemma [1] [2]. I suspect there's a better approach that I'm not seeing. Cheers, Rob [1]: see eg https://archive.nytimes.com/opinionator.blogs.nytimes.com/2010/06/20/the-anosognosics-dilemma-1/ (thanks Leo Butler) [2]: mentioned on-list here https://ffmpeg.org/pipermail/ffmpeg-user/2023-November/057183.html _______________________________________________ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
