*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

I suppose after you have downsized the image with pixel averaging you could just multiply all pixel values using ImageJ's Math functions. 

However, since the images are in RGB format, there is a risk that the intensity values when summed up will go over the 8-bit per channel limit, unless the original intensities were quite low. I didn't see an option when reducing image size to increase pixel depth (as when doing image arithmetic, for example). The option to increase bit depth does not seem to be applicable when doing simple math (such as multiplying the image by a factor of four, for instance). The only situation where bit depth can be increased is when doing image math (i.e. you could add one image to itself x number of times). However, I think in this situation, you would need to separate the RGB channels, do the math operations, and then recombine the resulting 16-bit R, G, B channels into a 48 bit color image. 

So not quite as simple, unless I am missing something....

One has also to wonder about what it means to "bin" a RGB image. With a Bayer mask camera, each group of four pixels actually represents two readings of green and one each of red and blue. The missing color values are interpolated. Binning in this situation is a bit more complex than when using a monochrome camera. This is why the method that would make most sense to me for an RGB image would be to separate the color channels, reduce the size of each channel by the appropriate "binning" factor (i.e. to 50%, 25%, etc, using pixel averaging. Remultiply by the same factor, increasing the bit depth if necessary to avoid saturation, and remerge channels into a color image (probably 48-bit). If you were using a monochrome camera to collect multi-channel fluorescent images, this method would be somewhat equivalent to actually binning the camera.

Now the question is, why do you want to do that, and is there an easier way to achieve the same goal?


--
Julio Vazquez
Fred Hutchinson Cancer Research Center
Seattle, WA 98109-1024


http://www.fhcrc.org



On Apr 26, 2011, at 8:13 AM, John Oreopoulos wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
> 
> Not exactly, 
> 
> Richard pointed out that a true binning would sum the pixels only, and not average them like the standard function does now. But I think the original ImageJ plugin or function could be modified to do this simply be removing the division part of the code. Richard, you could ask Wayne Rasband if he could include this option in the dialog box in a future version of ImageJ. 
> 
> John Oreopoulos
> Research Assistant
> Spectral Applied Research
> Richmond Hill, Ontario
> Canada
> www.spectral.ca
> 
> 
> On 2011-04-26, at 9:57 AM, Arne Seitz wrote:
> 
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>> 
>> Hi Rich,
>> 
>> your question has already been answered by John Oreopoulos yesterday.
>> 
>> But in fact you do not need a pluging. If you use ImageJ 1.44 and later you
>> can use Image-->Scale...
>> If you check "Average when downsizing" you get exactly what you were looking
>> for.
>> 
>> Cheers Arne