Posted by
Michael Schmid on
Feb 04, 2010; 3:53pm
URL: http://imagej.273.s1.nabble.com/RGB-Imaging-tp3689504p3689509.html
Hi,
of course, the histogram of the green channel just contains the
brightness of this channel. But if you compare, e.g., the histogram
of the green channel with that of the red channel, a more reddish
tone will give you am peak at higher values in the red channel and a
peak at lower values in the green one.
Similarily, the peak in the hue channel of a HSB stack will be
shifted towards red (i.e. towards 0 if the object is more greenish or
towards 255 if it is more bluish; in any case, more reddish means
further away from 128).
Michael
________________________________________________________________
On 4 Feb 2010, at 16:45, jspecs wrote:
> Michael,
>
>
>
>> just read the documentation to get a few ideas what could be done:
>>
>> Image>Color>Split Channels
>> Analyze>Histogram for each of the channels
>>
>> or
>>
>> Image>Type>HSB stack
>> Analyze>Histogram for each of the stack slices.
>> Especially the hue slice will be of interest to detect a change of
>> hue (see
http://en.wikipedia.org/wiki/HSV_color_space if you are not
>> familiar with HSB)
>
>
>
> I have tried what you mentioned, but both techniques seem to generate
> slices/channels with varying shades of black, gray, or white. This
> makes me
> think that the histogram will just generate its results based on
> brightness
> as opposed to RGB. Am I wrong?
>
>
>
> --
> View this message in context:
http://n2.nabble.com/RGB-Imaging-
> tp4511970p4514230.html
> Sent from the ImageJ mailing list archive at Nabble.com.