[Insight-users] Write new image registration metric

[Lydia Ng]

One more thought... each metric has three methods GetValue,
GetDerivative and GetValueAndDerivate. Different optimizers might use
a different method. Prehaps try to modify all three?

On 7/29/05, Dana Paquin <dpaquin at stanford.edu> wrote:
> Hi Lydia,
> 
> Thanks for your reply.  Yes, I watch the registration process at each
> step, and the printed metric values are exactly the same as before I
> added the *(0.25).  Please let me know if you have any ideas.
> 
> Thanks,
> Dana
> 
> Lydia Ng wrote:
> > Hi Dana,
> >
> > This seems a little strange - have you try adding some printout
> > statements to track what's going on?
> >
> > - Lydia
> >
> >
> > On 7/21/05, Dana Paquin <dpaquin at stanford.edu> wrote:
> >
> >>Hi all,
> >>
> >>I would like to know how to modify the existing ITK registration metrics
> >>to study the effect of a given metric on the registration process.  For
> >>example, suppose that I start with the itkMeanSquaresImageToImageMetric
> >>(as in ImageRegistration5.cxx).  To illustrate the process, suppose that
> >> I want the metric to be (1/4)*(mean squares).  In the file
> >>itkMeanSquaresImageToImageMetric.txx, I tried changing measure+=
> >>diff*diff to measure+=(diff*diff)*0.25, then saved this file and
> >>recompiled ImageRegistration5.  However, this had no effect on the
> >>metric values computed during the registration process.  Are there other
> >>files that I need to modify?  Can someone tell me what I need to do in
> >>order to modify a metric and then use that metric in the registration
> >>process?
> >>
> >>Thanks in advance,
> >>Dana
> >>
> >>Karthik Krishnan wrote:
> >>
> >>>The following earlier post might help:
> >>>http://public.kitware.com/pipermail/insight-users/2005-July/013868.html
> >>>http://public.kitware.com/pipermail/insight-users/2005-July/013866.html
> >>>
> >>>The last three paramters correspond to the translation, not the offset.
> >>>Take a look at the SoftwareGuide or the aforementioned mails to see the
> >>>difference.
> >>>
> >>>Thanks
> >>>regards
> >>>karthik
> >>>
> >>>Florence Dru wrote:
> >>>
> >>>
> >>>>hello,
> >>>>I'm running registration using the Euler3DTransform and I want to make
> >>>>   sure that I'm analyzing the right figures :
> >>>>       - if I want to get the real final offset I have to call the
> >>>>"GetOffset()" function don't I ? because if I only look at what I get
> >>>>   calling the "GetParameters()" function the last three parameters
> >>>>don't    really correspond to the final offset.
> >>>>   Thank you beforehand for confirming or correcting.
> >>>>
> >>>>Florence
> >>>>
> >>>>
> >>>>
> >>>>_______________________________________________
> >>>>Insight-users mailing list
> >>>>Insight-users at itk.org
> >>>>http://www.itk.org/mailman/listinfo/insight-users
> >>>>
> >>>
> >>>_______________________________________________
> >>>Insight-users mailing list
> >>>Insight-users at itk.org
> >>>http://www.itk.org/mailman/listinfo/insight-users
> >>>
> >>
> >>_______________________________________________
> >>Insight-users mailing list
> >>Insight-users at itk.org
> >>http://www.itk.org/mailman/listinfo/insight-users
> >>
>