Automatic location of spots at intersections of arcs of spots (Laue camera)

[Alan Hewat]

I am interested in the possibility of the automatic measurement of the
co-ordinates of isolated spots in an image, and eventually of spots
that lie at the intersection of arcs of spots, such as shown by the
numbered spots on http://www.neutronoptics.com/resources/oe-laue.jpg

This is a back-scattered Laue X-ray diffraction pattern, and is used
to determine the orientation of the crystal by solving for the (h,k,l)
indices of the diffraction spots; the spots at arc intersections
normally have low indices.
______________________________________________
Dr Alan Hewat, NeutronOptics, Grenoble, FRANCE
<[hidden email]> +33.476.98.41.68
        http://www.NeutronOptics.com/hewat
______________________________________________

[Michael Schmid]

Hi Alan,

use Process>Binary<Find Maxima to create an image with one point per  
maximum.
Then Analyze>Tool>Save XY Coordinates to export the list.

Finding the arcs automatically will be much more difficult.
Assuming that the arcs are ellipses, it would be possible if you  
write a Hough transform that takes advantage of the fact that all all  
ellipses have a vertex at the origin, so there are only 3 parameters  
for the ellipse (angle, a, b). Usually, a Hough transform with 3  
parameters should be possible (4 parameters would be too much).
(See the Wikipedia entry for Hough transform).

Hope this helps,

Michael
________________________________________________________________

On 17 Sep 2009, at 16:33, Alan Hewat wrote:

> I am interested in the possibility of the automatic measurement of the
> co-ordinates of isolated spots in an image, and eventually of spots
> that lie at the intersection of arcs of spots, such as shown by the
> numbered spots on http://www.neutronoptics.com/resources/oe-laue.jpg
>
> This is a back-scattered Laue X-ray diffraction pattern, and is used
> to determine the orientation of the crystal by solving for the (h,k,l)
> indices of the diffraction spots; the spots at arc intersections
> normally have low indices.
> ______________________________________________
> Dr Alan Hewat, NeutronOptics, Grenoble, FRANCE
> <[hidden email]> +33.476.98.41.68
>         http://www.NeutronOptics.com/hewat
> ______________________________________________

[Alan Hewat]

Thanks Michael. The arcs are indeed projections of ellipses. Before
some-one corrects my physics :-) my example was of forward scattering
not backscattering; with backscattering the arcs would be projections
of hyperbola. Alan.

2009/9/17 Michael Schmid <[hidden email]>:

> Hi Alan,
>
> use Process>Binary<Find Maxima to create an image with one point per
> maximum.
> Then Analyze>Tool>Save XY Coordinates to export the list.
>
> Finding the arcs automatically will be much more difficult.
> Assuming that the arcs are ellipses, it would be possible if you write a
> Hough transform that takes advantage of the fact that all all ellipses have
> a vertex at the origin, so there are only 3 parameters for the ellipse
> (angle, a, b). Usually, a Hough transform with 3 parameters should be
> possible (4 parameters would be too much).
> (See the Wikipedia entry for Hough transform).
>
> Hope this helps,
>
> Michael
> ________________________________________________________________
>
> On 17 Sep 2009, at 16:33, Alan Hewat wrote:
>
>> I am interested in the possibility of the automatic measurement of the
>> co-ordinates of isolated spots in an image, and eventually of spots
>> that lie at the intersection of arcs of spots, such as shown by the
>> numbered spots on http://www.neutronoptics.com/resources/oe-laue.jpg
>>
>> This is a back-scattered Laue X-ray diffraction pattern, and is used
>> to determine the orientation of the crystal by solving for the (h,k,l)
>> indices of the diffraction spots; the spots at arc intersections
>> normally have low indices.

______________________________________________
Dr Alan Hewat, NeutronOptics, Grenoble, FRANCE
<[hidden email]> +33.476.98.41.68
        http://www.NeutronOptics.com/hewat
______________________________________________

[David Webster]

Alan,

Michael's idea about using a Hough Transform is intriging to an old
Hough'ophile like me. I looked for reference's on this subject but only
found ones that used edge location and direction. If you find one that
using point location only, I would be interested in it.

David Webster
[hidden email]

[dpoburko]

I have modified the current verison of the JACOP plugin so that it sends
a correlatoin coefficient to the results table when  it runs a Pearson
Correlation. Then I run a simple macro that runs a Pearson Correlation
after translating one image with respect to another over an abritrary
size grid. The location of best fit is recorded, and essentially you get
a the x,y translation of the original image with respect to a reference
image. It's a rather slow, ugly hack, but it gets the job done. I'm
happy to share it if it can be of help.

However, this was just a poor man's work around. I agree with some other
comments that it's surprising that other plugins like TurboReg don't
make this kind of info easier to get at. That would be super useful.

Cheers,
Damon



Alan Hewat wrote:

> I am interested in the possibility of the automatic measurement of the
> co-ordinates of isolated spots in an image, and eventually of spots
> that lie at the intersection of arcs of spots, such as shown by the
> numbered spots on http://www.neutronoptics.com/resources/oe-laue.jpg
>
> This is a back-scattered Laue X-ray diffraction pattern, and is used
> to determine the orientation of the crystal by solving for the (h,k,l)
> indices of the diffraction spots; the spots at arc intersections
> normally have low indices.
> ______________________________________________
> Dr Alan Hewat, NeutronOptics, Grenoble, FRANCE
> <[hidden email]> +33.476.98.41.68
>         http://www.NeutronOptics.com/hewat
> ______________________________________________
>  

--

Damon Poburko, PhD
Postdoctoral Research Fellow
Stanford University School of Medicine
Dept. of Molecular & Cellular Physiology
279 Campus Dr., Beckman B103, Stanford, CA 94305
Ph: 650 725 7564, fax: [phone]