Re: Similarity between integers in a list
Re: Similarity between integers in a list
- Subject: Re: Similarity between integers in a list
- From: Christopher Nebel <email@hidden>
- Date: Tue, 18 May 2004 18:54:29 -0700
On May 18, 2004, at 4:01 PM, Martin Orpen wrote:
I'm working on a Photoshop script that tries to work out whether an
RGB image is colourful or neutral.
The script gets RGB values from 6 (or more) randomly chosen pixels in
the form:
Image 1
{{71, 94, 97}, {95, 96, 98}, {60, 68, 73}, {41, 49, 56}, {21, 23, 36},
{44, 47, 52}}
Image 2
{{89, 89, 89}, {82, 82, 82}, {77, 77, 77}, {48, 49, 49}, {48, 49, 50},
{66, 68, 68}}
I'd like to process each list so that I can differentiate between the
images - image 1 being a conventional colour image where the lists
show greater variation and image 2 being neutral and the pixel values
tend to be nearly equal.
It's not really clear what you mean by "colourful". Do you mean that
there are a lot of different colors? That the colors are all well away
from 50% gray? That there are a lot of high-saturation-high-brightness
(for lack of a better word, "colourful") colors? For instance, a
black-to-white wipe would qualify as "colourful" under the first and
second definitions, but not the third. A pure red image would qualify
under the second and third definitions, but not the first. A pure
black image would qualify under the first definition, but not the
second or third. You see the problem?
Also, you've currently got everything expressed as RGB, but distance in
RGB space isn't good at determining all color qualities. L*a*b* would
make more sense for the first two definitions; HSB would make more
sense for the third. (I'd be surprised if Photoshop couldn't do the
conversion for you.)
--Chris Nebel
AppleScript Engineering
_______________________________________________
applescript-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/applescript-users
Do not post admin requests to the list. They will be ignored.