[Insight-users] Combining 3 images

[Luis Ibanez]

Hi Vera,

Thanks for letting us know about the details of your work.

You may find interesting to know that we recently added
a Curve3DExtractor application to InsightApplications.

http://www.itk.org/cgi-bin/viewcvs.cgi/Curves3DExtractor/?root=InsightApplications

You may want to update your CVS checkout of InsightApplications
and give it a try.

There is still some work pending in this application,
but we anticipate it to be usable in a matter of days.

The current version of the application computes the Hessian
of the input image, as well as teh Gradient. The eigenvalues
of the Hessian are "almost" avaialble, ... as soon as we
figure out some issues with memory allocation.

We will be very interested in having your feedback on this
application.


     Best regards,



        Luis


----------------------------------
vera-naumburger at freenet.de wrote:
> Hi Luis,
> 
> great, thanks a lot. In this special case I'm not working with Second Derivatives.
> Sorry, I wanted to combine the First Derivatives ( Gradient x, Gradient y and Gradient z).
> 
> Actually I´m working on one of the examples provided with ITK. I´m trying to convert the 2D Curves Extractor into a 3D Curves Extractor. I´m reading CT images and let them perform all the actions known from the 2D Curves Extractor.
> I'm quite suprised that most of the converting is quite comfortable :-)))
> 
> My question was aiming at the scalar product of the GradientVectorImage with one of the EigenVektors. So I wasn`t sure which 
> filter to use instead of the JoinImageFilter to get the GradientVectorImage. But your suggestions sound good to me.
> I was trying to use the GradientImageFilter on the xyz-Smothed Image. But result with the scalar product looks a little weird.
> I going to try your version.
> 
> So might be, that I have some othe questions soon :-)
> 
> Thank you a lot!
> Vera
> 
> ----- original Nachricht --------
> 
> Betreff: Re: [Insight-users] Combining 3 images
> Gesendet: Di 24 Mai 2005 17:32:07 CEST
> Von: "Luis Ibanez"<luis.ibanez at kitware.com>
> 
>>
>>Hi Vera,
>>
>>
>>
>>The filter that you need is the
>>
>>
>>itkCompose3DVectorImageFilter
>>http://www.itk.org/Insight/Doxygen/html/classitk_1_1Compose3DVectorImageFilt
>>er.html
>>
>>
>>and/or the
>>
>>
>>itkCompose3DCovariantVectorImageFilter
>>http://www.itk.org/Insight/Doxygen/html/classitk_1_1Compose3DCovariantVector
>>ImageFilter.html
>>
>>
>>
>>Note that ITK differentiates between "Vectors" which are defined
>>as the difference between two positions in space, and "CovariantVectors"
>>which are defined as the orthogonal direction to a (N-1)-dimensional
>>subspace.  Covariant vectors represent Gradients, and Normals to
>>Surfaces (in 3D), while "Vectors" represent the relative position
>>between two points.
>>
>>
>>This distinction is quite important when you apply AffineTransforms
>>to this geometrical objects.
>>
>>
>>
>>Now that... from your email, you seems to be dealing with Second
>>Derivatives, in which case you are dealing with a symmetric Tensor
>>of Second Rank with two Covariant components. In that context you
>>should be careful since this object does  not behave as a Vector,
>>nor a CovariantVector.  In case you need it, we recently add a
>>Second rank symmetric tensor class to ITK.
>>
>>
>>
>>Please let us know if you have any further questions,
>>
>>
>>
>>     Thanks
>>
>>
>>
>>
>>       Luis
>>
>>
>>
>>----------------------
>>Vera Naumburger wrote:
>>
>>
>>>Hi,
>>>
>>>I´m trying to combine three gradient images (xy, yz, zx), so that the  
>>>output is a VectorImage. I found the JoinImageFilter for
>>>joining 2 images, is there something like that for 3 images, as well?
>>>
>>>Thanks,
>>>Vera
>>>_______________________________________________
>>>Insight-users mailing list
>>>Insight-users at itk.org
>>>http://www.itk.org/mailman/listinfo/insight-users
>>>
>>>
>>
>>
>>
>>
> 
> --- original Nachricht Ende ----
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
>