MJPEG Tools Svn

*** empty log message ***

Rewrote the top-level denoising. The search-border is no longer used.
For every pixel-group, the denoiser now keeps the match-count-throttle
matches with the lowest SAD, and flood-fills each until the resulting
region is at least match-size-throttle pixel-groups big, then applies
that region right then and there. This new algorithm is a lot simpler
than the previous one, but seems to work a lot better.

Also, a nasty bug was fixed in the way the extent of applied regions
are removed from the search-window cells. VERY nasty bug. It
technically produced broken data.

Image quality has improved DRAMATICALLY. Performance has increased from
anywhere between 20% (for normal video) to 35% (for pathological video).

Purge() now takes an x/y offset, to allow moved regions to be processed
correctly.

Now purges the pixel-sorter tree every 10 frames, instead of every 100
frames. Changing algorithms require changing fudge factors...

Changed the default match-size throttle (since it means something
different now).

partly implemented general-sampling-theorem interpolation.
The image-quality now is far far better. memory handly is
a remaining problem, as some functions need to overshoot for
reasons of speed...
So it may not work on MACs/PPC... :((

Need to add the $extra_cflags to CXXFLAGS for the C++ modules (mpeg2enc and mplex)

min = 2 * abs(*(frame+x+(y-1)*width) - *(frame+x+(y+1)*width)) is a lot
better (especially inside a double nested loop) than:
d = *(frame+x+(y-1)*width) - *(frame+x+(y+1)*width) ;
min = d < 0 ? -d:d;
d = *(frame+x+(y-1)*width) - *(frame+x+(y+1)*width) ;
min += d < 0 ? -d:d;
Also do a bit of tidying (unused/unneded ABS macro removed)

save a few cpu cycles, abs() is faster than (xxx < 0 ? -xxx : xxx)

Adjust the CFLAGS/AM_CFLAGS to use the new new PROGRAM_NOPIC option. Extra
flags (the old AM_CFLAGS) should be added by "--with-extra-cflags=..." at
./configure time (so that they apply to all programs and not just yuvdenoise)