[MPlayer-dev-eng] [CFTWWI] VF logodetect/logoremove

[Ivo]

On Thursday 11 August 2005 18:26, Rich Felker wrote:
> Of course it's probably shitty code that uses doubles all over the
> place, and all sorts of other inefficiencies. Just rewriting it
> properly would probably get a 10x speedup, without even considering
> algorithmic optimizations. Of course this is just a guess, based on
> the way academic proof of concept code usually works..
[..]
> Optimized, this would be perfectly fine. I'm used to 1fps or slower
> encoding anyway.

Yes. The code uses a C++ library which is basically one big (550kB) .h file 
and at times barely readable. Datatypes are overly-complex, loops are the 
wrong way around (for x as outer loop, etc.) and all calculations are done 
in floats.

Right now I'm trying to understand the code and the paper and write some 
testcode in C. Maybe I'll turn that into a filter later. I'm not sure how 
it will perform on moving images though. The basic algorithm could be 
expanded to include multiple images at once, but for now, I'm having enough 
trouble understanding it as it is :)

--Ivo