grid aligner plugin error

[Jeff Spector-2]

when attempting to run the "GridAligner" Plugin. I get the following error :

 Class not found while attempting to run "GridAligner_"
java.lang.NoClassDefFoundError:org/jdesktop/layout/GroupLayout$Group


I've followed the instructions here
http://valelab.ucsf.edu/~nstuurman/ijplugins/GridAligner.html

I'm using Java 1.6.0_65 on a mac with OS 10.8.5

can someone help?
Thanks..
-Jeff

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Nico Stuurman]

Hi Jeff,

> when attempting to run the "GridAligner" Plugin. I get the following error :
>
>   Class not found while attempting to run "GridAligner_"
> java.lang.NoClassDefFoundError:org/jdesktop/layout/GroupLayout$Group
>
>
> I've followed the instructions here
> http://valelab.ucsf.edu/~nstuurman/ijplugins/GridAligner.html
Did you try downloading
http://valelab.ucsf.edu/~nstuurman/ijplugins/download/plugins/GridAligner/swing-layout-1.0.4.jar 
and copying it to your plugins folder?  Even though you are running Java
1.6, it seems to be missing from the classpath.

Best,

Nico

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Jeff Spector-2]

On Thu, Dec 10, 2015 at 1:08 PM, Jeff Spector <[hidden email]> wrote:

>
>
> On Wed, Dec 9, 2015 at 10:55 PM, Nico Stuurman <[hidden email]>
> wrote:
>
>> Hi Jeff,
>>
>> when attempting to run the "GridAligner" Plugin. I get the following
>>> error :
>>>
>>>   Class not found while attempting to run "GridAligner_"
>>> java.lang.NoClassDefFoundError:org/jdesktop/layout/GroupLayout$Group
>>>
>>>
>>> I've followed the instructions here
>>> http://valelab.ucsf.edu/~nstuurman/ijplugins/GridAligner.html
>>>
>> Did you try downloading
>> http://valelab.ucsf.edu/~nstuurman/ijplugins/download/plugins/GridAligner/swing-layout-1.0.4.jar
>> and copying it to your plugins folder?  Even though you are running Java
>> 1.6, it seems to be missing from the classpath.
>>
>> Best,
>>
>> Nico
>>
>> --
>> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>>
>
>
> Hi,
>  Yes I did that, the swing-layout-1.0.4.jar is in the plugins folder and I
> still get the error. Any other suggestions ?
>  Thanks..
> -Jeff
>
>

I've got grid aligner working now ( I just removed it and reinstalled it
under FIJI and it works now.
The alignment between my images looks perfect, but it looks like the plugin
is changing my intensity values.
 Is this normal?  as an example I drew  a line in the raw data and the
aligned data in the aligned data the min and max are 0 and 257 where as in
the original raw data they are 250.69 and 386.89
I noticed on the website that a bug was corrected that gave weird intensity
values. Can someone please verify to me that it is indeed not changing my
intensity values? the grid I used to align my two channels is much brighter
than my images are but I don't think that should matter.
 Can someone please help and/or explain to me why the intensity values are
different in the  (perfectly) aligned images?
 thanks!
-Jeff

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Nico Stuurman]

Hi Jeff,

On 12/10/15 2:29 PM, Jeff Spector wrote:
> I've got grid aligner working now ( I just removed it and reinstalled it
> under FIJI and it works now.
> The alignment between my images looks perfect, but it looks like the plugin
> is changing my intensity values.
>   Is this normal?  as an example I drew  a line in the raw data and the
> aligned data in the aligned data the min and max are 0 and 257 where as in
> the original raw data they are 250.69 and 386.89

Looks like you are working with images of type float (32-bit images).  I
never tried those, and certainly did not provide a code-path for them,
so it is unclear what will happen.  You could try converting first to
16-bit images.



Best,

Nico

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Jeff Spector-2]

On Fri, Dec 11, 2015 at 12:24 AM, Nico Stuurman <[hidden email]>
wrote:

> Hi Jeff,
>
> On 12/10/15 2:29 PM, Jeff Spector wrote:
>
>> I've got grid aligner working now ( I just removed it and reinstalled it
>> under FIJI and it works now.
>> The alignment between my images looks perfect, but it looks like the
>> plugin
>> is changing my intensity values.
>>   Is this normal?  as an example I drew  a line in the raw data and the
>> aligned data in the aligned data the min and max are 0 and 257 where as in
>> the original raw data they are 250.69 and 386.89
>>
>
> Looks like you are working with images of type float (32-bit images).  I
> never tried those, and certainly did not provide a code-path for them, so
> it is unclear what will happen.  You could try converting first to 16-bit
> images.
>
>
>
>
> Best,
>
> Nico
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

Hi,
 The images I'm using are 16-bit (they come from an andor ixon camera) What
I am seeing is that after calculating the transform when I apply it the new
image is registered fine, but when I look at the image histograms, the
histogram for channel 2 is always a series of discrete spikes, where as in
the original image it was a smooth looking histogram. The histogram for
channel 1 remains unchanged. Please let me know what you think..
-Jeff

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Nico Stuurman]

> I've got grid aligner working now ( I just removed it and reinstalled it
> under FIJI and it works now.
> The alignment between my images looks perfect, but it looks like the
> plugin
> is changing my intensity values.
>    Is this normal?  as an example I drew  a line in the raw data and the
> aligned data in the aligned data the min and max are 0 and 257 where as in
> the original raw data they are 250.69 and 386.89
>
>> Looks like you are working with images of type float (32-bit images).  I
>> never tried those, and certainly did not provide a code-path for them, so
>> it is unclear what will happen.  You could try converting first to 16-bit
>> images.
>   The images I'm using are 16-bit (they come from an andor ixon camera) What
> I am seeing is that after calculating the transform when I apply it the new
> image is registered fine, but when I look at the image histograms, the
> histogram for channel 2 is always a series of discrete spikes, where as in
> the original image it was a smooth looking histogram. The histogram for
> channel 1 remains unchanged. Please let me know what you think..

Hmm.  I can indeed reproduce this odd behavior.  I seem to be getting
discrete values spaced at 256 intervals (as if the affine transform
takes place in 8 bit and the result is scaled back to 16-bit).  The code
uses the java class java.awt.image.AffineTransformOp to execute the
affine transform. That class has a setting for the method of "filling
in" pixel values.  By changing the method from Bicubic to Bilinear, the
behavior improves, and I no longer see discrete values in the output.  I
made that changed code available as version 0.13.

Eventually, it would be nice for the plugin to be transformed into a
Fiji plugin and to use imglib2 (or Image Ops) to execute the affine
transform.  Maybe when I retire... (or if anyone else feels up to the
task, code is available here:
https://valelab.ucsf.edu/svn/valelabtools/IJPlugins/GridAligner/src/).

Best,

Nico

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html