[ITK-dev] [ITK] 2D/3D registration in v4

[David Burns]

I think this issue relates to the way the metrics compute gradients. It 
can be reproduced using

Examples/RegistrationITKv4/ImageRegistration1.cxx
and
Examples/RegistrationITKv3/ImageRegistration1.cxx

by setting Dimensions to 3, and using 2D images as inputs.

I found using: metric->ComputeGradientOff()
will allow the registration to proceed for ITKv3 using the 
MeanSquaresImageToImageMetric, but there is no such function for the v4 
version.
Using a metric that does not require gradients (eg Mattes) will allow 
the registration to proceed in both v3 and v4.

I humbly suggest the registration framework and metrics should be able 
to handle this situation. Also of note to those interested, the v3 
registration on my system goes about twice as fast per iteration.

What are your thoughts?

David

On 11/12/2015 08:41 PM, Matias Montroull wrote:
>
> David, would you be able to share the code you wrote?
>
> On Nov 12, 2015 10:39 PM, "David Burns" <david.mo.burns at gmail.com 
> <mailto:david.mo.burns at gmail.com>> wrote:
>
>     Dear Insight,
>
>     I am having the following trouble attempting 2D/3D registration of
>     xray/CT with ITK v4 registration framework.
>
>     The basics: The moving image is a 3D volume, and the fixed image
>     is a 2D projection. An interpolator calculates a synthetic
>     cone-beam projection for the moving image (volume) and compares to
>     the fixed 2D image.
>
>     The v4 registration framework doesn't allow me to use a 2D fixed
>     and 3D moving image -> won't compile. So I need to setup the fixed
>     image as 3D, even though it is 2D data. I can compile, but I get
>     the following error at run time
>
>
>     ExceptionObject caught !
>
>     itk::ExceptionObject (0x20cfac0)
>     Location: "void itk::RecursiveSeparableImageFilter<TInputImage,
>     TOutputImage>::BeforeThreadedGenerateData() [with TInputImage =
>     itk::Image<float, 3u>; TOutputImage = itk::Image<float, 3u>]"
>     File: /usr/local/include/ITK-4.8/itkRecursiveSeparableImageFilter.hxx
>     Line: 245
>     Description: itk::ERROR: RecursiveGaussianImageFilter(0x2089fb0):
>     The number of pixels along direction 2 is less than 4. This filter
>     requires a minimum of four pixels along the dimension to be processed.
>
>     It seems this behavior does not occur in the v3 registration
>     framework and 2D/3D registrations were possible, however, I want
>     to take advantage of the threading features in v4. How would you
>     suggest I proceed?
>
>
>     Thanks
>     David Burns
>
>
>
>
>
>
>     _______________________________________________
>     Powered by www.kitware.com <http://www.kitware.com>
>
>     Visit other Kitware open-source projects at
>     http://www.kitware.com/opensource/opensource.html
>
>     Kitware offers ITK Training Courses, for more information visit:
>     http://kitware.com/products/protraining.php
>
>     Please keep messages on-topic and check the ITK FAQ at:
>     http://www.itk.org/Wiki/ITK_FAQ
>
>     Follow this link to subscribe/unsubscribe:
>     http://public.kitware.com/mailman/listinfo/insight-developers
>     _______________________________________________
>     Community mailing list
>     Community at itk.org <mailto:Community at itk.org>
>     http://public.kitware.com/mailman/listinfo/community
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/insight-developers/attachments/20151112/a36a2532/attachment.html>