> Is there EXIF info in the image?
EXIF doesn't give any image information, only meta information (shooting
parameters, date time, etc).
In principle, the problem looks to me like it wants some statistical
function run on the pixel brightness values. That gives two issues:
define "statistical function", though some simple average will probably
do you, and how to do this will minimal programming and
non-interactively. Supposedly gimp can be totally scripted, but I
believe it when I see it (and chances are they used some archaic mental
thing like scheme).
> >I am thinking of the imagemagick 'identify' command, and maybe extracting
> >the mean value of the red, green and blue channels and comparing them to a
> >hard limit.
Probably the solution with the least amount of programming and fuss,
although heavy on the CPU. Next step up would be to use the ImageMagick
perls interface (arrgh) or the python one, though my impression was the
latter lacks in development.
> is there a better tool?
I don't know of any image toolbox which presents an easy interface for
custom functions, though it would be handy to have. I can see these
alternatives:
1) Hack into ImageMagick and add your own function; perhaps you're lucky
and can adopt an existing one.
2) Check if octave can read jpeg data. Keep in mind that netpbm can
translate most formats into pnm, which is a trivially simple binary
format often used for image interchange. Octave might read that if it
doesn't read jpeg directly. Octave can be trivially scripted with
octave < your-fancy-octave-script.oct
Volker
--
Volker Kuhlmann is list0570 with the domain in header
http://volker.dnsalias.net/ Please do not CC list postings to me.
