resolution affect intensity?

[shilo2]

Dear list members
Maybe it’s a stupid question but because I'm a beginner I must ask  
In a mistake I took some pictures (confocal microscope) in low resolution (256*256). The majority are in more high resolution (512*512). Now, I would like to compare intensity's.
Does the resolution should affect the intensity means? If the answer is yes , is there a way to correct it?

Thank you  

[Ben Tupper]

On Dec 8, 2008, at 7:56 AM, shilo2 wrote:

> Dear list members
> Maybe it’s a stupid question but because I'm a beginner I must ask
> In a mistake I took some pictures (confocal microscope) in low  
> resolution
> (256*256). The majority are in more high resolution (512*512). Now,  
> I would
> like to compare intensity's.
> Does the resolution should affect the intensity means? If the  
> answer is yes
> , is there a way to correct it?
>

Hi,

That doesn't seem like a stupid question to me.  I assume you mean  
greyscale intensity - in which case your best way to make the  
comparison is to "normalize" the integrated density by area.  This  
gives you the mean grey value which is one of ImageJ's measurement  
selections (see http://rsb.info.nih.gov/ij/docs/menus/analyze.html#set).

normalized grayness = mean grey value = (integrated density) / area

Once normalized, you can reasonably compare across different  
resolutions.*
Hope that helps.

Cheers,
Ben

* There are other issues involved, including how the pixel binning  
was done between the low and high resolution.  If you need to dig  
into that you'll have to consult your camera documentation and an  
image processing reference. This one (http://www.imagingbook.com/) is  
a goodie.

[David Hovis]

It certainly can, depending on your microscope.  Changing your  
resolution can change your raster speed or your time/pixel.  Some  
instruments will attempt to auto-adjust the PMT voltages when you make  
a change.

The only way to find how big the effect might be would be to take  
sequential images at 256 and 512 resolutions (on a sample that won't  
photobleach during scans) and compare.

--David


On Dec 8, 2008, at 7:56 AM, shilo2 wrote:

> Dear list members
> Maybe it’s a stupid question but because I'm a beginner I must ask
> In a mistake I took some pictures (confocal microscope) in low  
> resolution
> (256*256). The majority are in more high resolution (512*512). Now,  
> I would
> like to compare intensity's.
> Does the resolution should affect the intensity means? If the answer  
> is yes
> , is there a way to correct it?
>
> Thank you
>
>
> --
> View this message in context: http://n2.nabble.com/resolution-affect-intensity--tp1628924p1628924.html
> Sent from the ImageJ mailing list archive at Nabble.com.

[karo03]

Typically cl(sm) parameters are changed, changing the resolution or  
optics. I would think mean intensities of regions are not directly  
comparable. At least a different gain factor is to be expected. I  
would calculate that gain factor by measuring similar objects which  
should have the same (mean) intensity. The additive offset might be  
zero for fluerescent images.
Regards
Karsten

Am 08.12.2008 um 13:56 schrieb shilo2:

> Dear list members
> Maybe it’s a stupid question but because I'm a beginner I must ask
> In a mistake I took some pictures (confocal microscope) in low  
> resolution
> (256*256). The majority are in more high resolution (512*512). Now,  
> I would
> like to compare intensity's.
> Does the resolution should affect the intensity means? If the answer  
> is yes
> , is there a way to correct it?
>
> Thank you
>
>
> --
> View this message in context: http://n2.nabble.com/resolution-affect-intensity--tp1628924p1628924.html
> Sent from the ImageJ mailing list archive at Nabble.com.

[shilo2]

Hi, Ben

thanks for your answer

 

I have used the Multi Measure in the ROI manager.

If I understand you correctly this should give me the average - intensity/area and this is already normalized data

 

Shilo

On Mon, Dec 8, 2008 at 4:26 PM, Ben Tupper (via Nabble) <[hidden email]> wrote:

On Dec 8, 2008, at 7:56 AM, shilo2 wrote:

> Dear list members
> Maybe it's a stupid question but because I'm a beginner I must ask
> In a mistake I took some pictures (confocal microscope) in low  
> resolution
> (256*256). The majority are in more high resolution (512*512). Now,  
> I would
> like to compare intensity's.
> Does the resolution should affect the intensity means? If the  
> answer is yes
> , is there a way to correct it?
>

Hi,

That doesn't seem like a stupid question to me.  I assume you mean  
greyscale intensity - in which case your best way to make the  
comparison is to "normalize" the integrated density by area.  This  
gives you the mean grey value which is one of ImageJ's measurement  
selections (see http://rsb.info.nih.gov/ij/docs/menus/analyze.html#set).

normalized grayness = mean grey value = (integrated density) / area

Once normalized, you can reasonably compare across different  
resolutions.*
Hope that helps.

Cheers,
Ben

* There are other issues involved, including how the pixel binning  
was done between the low and high resolution.  If you need to dig  
into that you'll have to consult your camera documentation and an  
image processing reference. This one (http://www.imagingbook.com/) is  
a goodie.

---

This email is a reply to your post @ http://n2.nabble.com/resolution-affect-intensity--tp1628924p1629209.html
You can reply by email or by visting the link above.

[Gabriel Landini]

On Monday 08 December 2008 14:40:54 shilo2 wrote:
> I have used the Multi Measure in the ROI manager.
>
> If I understand you correctly this should give me the average -
> intensity/area and this is already normalized data

You most likely changed various settings in the microscope to get a reasonable
image in both cases, so image intensity is unlikely to be directly comparable
across magnifications.

Cheers,
G.

[Michael Doube]

I agree with Gabriel; yes the intensities will be different. Unless you
know everything about your confocal you will not be able to compare
brightnesses across images of different resolutions.  It is hard enough
when you do your best to keep everything exactly the same...  For
example, the Leica SP5 automagically adjusts for resolution effects on
brightness but the earlier Leica SP2 does not.  So you must know what
the automagic is before trying to do any correction to your broken
images.  Much better is just to curse your mistake and do the imaging again.

Mike

Gabriel Landini wrote:

> On Monday 08 December 2008 14:40:54 shilo2 wrote:
>> I have used the Multi Measure in the ROI manager.
>>
>> If I understand you correctly this should give me the average -
>> intensity/area and this is already normalized data
>
> You most likely changed various settings in the microscope to get a reasonable
> image in both cases, so image intensity is unlikely to be directly comparable
> across magnifications.
>
> Cheers,
> G.

[shilo2]

I udersatnd, no shortcuts
 Thank you all

Shilo



On Mon, Dec 8, 2008 at 5:15 PM, Michael Doube <[hidden email]>wrote:

> I agree with Gabriel; yes the intensities will be different. Unless you
> know everything about your confocal you will not be able to compare
> brightnesses across images of different resolutions.  It is hard enough when
> you do your best to keep everything exactly the same...  For example, the
> Leica SP5 automagically adjusts for resolution effects on brightness but the
> earlier Leica SP2 does not.  So you must know what the automagic is before
> trying to do any correction to your broken images.  Much better is just to
> curse your mistake and do the imaging again.
>
> Mike
>
>
> Gabriel Landini wrote:
>
>> On Monday 08 December 2008 14:40:54 shilo2 wrote:
>>
>>> I have used the Multi Measure in the ROI manager.
>>>
>>> If I understand you correctly this should give me the average -
>>> intensity/area and this is already normalized data
>>>
>>
>> You most likely changed various settings in the microscope to get a
>> reasonable image in both cases, so image intensity is unlikely to be
>> directly comparable across magnifications.
>>
>> Cheers,
>> G.
>>
>