Periodicity

[Bennett Walker]

Dear Image J help,

 

I have recently begun using image j for the analysis of some SEM images that
I have taken of an array of columns. I can visually see that there is some
sort of periodicity among my columns, however; when I use the FFT feature
from image j I can not obtain any usable information. The purpose of using
this program is to determine differences in periodicity between numerous
different SEM images similar to this one.

 

Below I have a link to access an example of the images and the FFT of the
image.

 

Do you have any suggestions about how I can determine periodicity with these
SEM images?

 

Regards,

 

-Bennett Walker

 

 

http://home.gwu.edu/~bennettw/FFT%20of%20image.jpg

 

http://home.gwu.edu/~bennettw/For%20Image%20J.jpg

 

 

[Unruh, Jay]

Bennett,

Try FD Math with operation correlate and do inverse transform selected.  This will give you a 2D correlation that corresponds approximately to the spatial probability distribution.  The size of the main center peak will be related to the average size of your objects and the ring surrounding it relates to the average distance between objects.

Jay Unruh
Research Specialist
Stowers Institute for Medical Research

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Bennett Walker
Sent: Tuesday, June 10, 2008 11:36 AM
To: [hidden email]
Subject: Periodicity

Dear Image J help,



I have recently begun using image j for the analysis of some SEM images that I have taken of an array of columns. I can visually see that there is some sort of periodicity among my columns, however; when I use the FFT feature from image j I can not obtain any usable information. The purpose of using this program is to determine differences in periodicity between numerous different SEM images similar to this one.



Below I have a link to access an example of the images and the FFT of the image.



Do you have any suggestions about how I can determine periodicity with these SEM images?



Regards,



-Bennett Walker





http://home.gwu.edu/~bennettw/FFT%20of%20image.jpg



http://home.gwu.edu/~bennettw/For%20Image%20J.jpg

[Charles M. Shaw]

It appears to me that you don't so much have a periodicity in the
distribution of these columns, but rather a fairly consistent
inter-column spacing.  The rings you see in your FFT would then be a
statistical measure, of sorts, of the inter-column distance within the
image (much like diffraction from an isotropic liquid or glass) and the
size (i.e. diameter) of the columns themselves.

Charlie Shaw
Graduate Student
Macromolecular Science & Engineering Program
University of Michigan, Ann Arbor

On Jun 10, 2008, at 12:35 PM, Bennett Walker wrote:

> Dear Image J help,
>
>
>
> I have recently begun using image j for the analysis of some SEM
> images that
> I have taken of an array of columns. I can visually see that there is
> some
> sort of periodicity among my columns, however; when I use the FFT
> feature
> from image j I can not obtain any usable information. The purpose of
> using
> this program is to determine differences in periodicity between
> numerous
> different SEM images similar to this one.
>
>
>
> Below I have a link to access an example of the images and the FFT of
> the
> image.
>
>
>
> Do you have any suggestions about how I can determine periodicity with
> these
> SEM images?
>
>
>
> Regards,
>
>
>
> -Bennett Walker
>
>
>
>
>
> http://home.gwu.edu/~bennettw/FFT%20of%20image.jpg
>
>
>
> http://home.gwu.edu/~bennettw/For%20Image%20J.jpg
>
>
>
>
>

[Robert Dougherty]

Bennett,

The picture does not look very periodic to me, but it does seem to  
have short range order, if that is the right term.  Perhaps you could  
could do a particle analysis to find the centers of mass of the  
spots, and then make a histogram of the radial distances between the  
centers of mass.  (Compute the full matrix.)  The first big peak in  
this histogram would be a preferred distance. It looks like it should  
be about 40 pixels.

Bob

On Jun 10, 2008, at 9:35 AM, Bennett Walker wrote:

> Dear Image J help,
>
>
>
> I have recently begun using image j for the analysis of some SEM  
> images that
> I have taken of an array of columns. I can visually see that there  
> is some
> sort of periodicity among my columns, however; when I use the FFT  
> feature
> from image j I can not obtain any usable information. The purpose  
> of using
> this program is to determine differences in periodicity between  
> numerous
> different SEM images similar to this one.
>
>
>
> Below I have a link to access an example of the images and the FFT  
> of the
> image.
>
>
>
> Do you have any suggestions about how I can determine periodicity  
> with these
> SEM images?
>
>
>
> Regards,
>
>
>
> -Bennett Walker
>
>
>
>
>
> http://home.gwu.edu/~bennettw/FFT%20of%20image.jpg
>
>
>
> http://home.gwu.edu/~bennettw/For%20Image%20J.jpg
>
>
>
>
>

Robert Dougherty, Ph.D.
President, OptiNav, Inc.
10900 NE 8th St, Suite 900
Bellevue, WA 98004
Tel. [phone]
FAX [phone]
www.optinav.com
[hidden email]