Re[2]: [XviD-devel] Global Motion Compensation

[Radek 'sysKin' Czyz]

------------9E1851103F01E877
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

> let's combine our effords in this, because I had similar plans:
> I guess fast GMC has to work like that: First calculate global
> translation (e.g. from motion vectors or another way), then refine to
> more free parameters (sheer/zoom)...

I haven't thought beyond one-point GMC yet.

> One crucial step is _which_ motion vectors from ME to consider in getting
> translation and which not. There are several ways for that, maybe we can
> compare which of our ideas works best before switching to the next step.

Now this is what I'm pround of ;)
My reasoning was: the most popular ones, while ignoring the others.
How to find them? I thought about some kind of a histogram, but I
found a better way.
I'll attach the function here. Call it after ME to see how it works -
or before ME if max_bframes >=0 OR after reading hinted ME data.

> Do you have any fixed test clips in mind such we can compare results?
> I guess the Matrix Lobby scene isn't too bad, agreed, but we should have
> other material, too.

Well I made myself some clips (from hq divx, resized, stored in
lossless codec) and I'm using them for all my tests...

> It may be more fun in the beginning, but it's also dangerous because you
> might follow tracks that were proven to be bad. So if you want to have
> one or two good general papers, just tell me.
> I won't tell you the URLs right now so you have a choice to say "no" ;-)

No, I'm not saying 'no'. Sure, gemme URLs.

Radek

PS Ah right, the code. attaching now
Oh, I just remembered - P/B/I decision needs to be fixed a bit to work
with it (pMB-> mode needs to be set to INTER if it found a vector, or
INTRA if deviation was bigger then SAD. It wasn't important for the
function before).
Ah, ask me what the code does (ie how) 'cause I really want to send
this mail NOW ;)
------------9E1851103F01E877
Content-Type: text/plain; name="gmc.txt"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="gmc.txt"
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=

------------9E1851103F01E877--