Hi Tim,
On Mon, 25 Feb 2013, Tim Bogatchev wrote:
> Hello, I am currently doing a project involving print verification, and
> I was hoping to use the ImageJ library to compare two images and
> determine whether they are identical (or almost identical).
>
> Is ImageJ a good library to do this with and if so what should my
> general strategy be?
I would think so. It would not hurt to get acquainted with color space
theory, but as a quick solution I would suggest to use one of the color
space converters to convert the RGB image to CIELAB, then use the Image
Calculator to calculate the difference, and then use Analyze>Measure
(possibly after Analyze>Set Measurements... to select the appropriate
values) to calculate mean, median and maximum.
Of course, this assumes that your original images are encoded in a known
RGB (such as sRGB) and not hampered by Jpeg artifacts. As I said, getting
into color space theory a bit would not hurt a bit ;-)
Ciao,
Johannes
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
