[Insight-users] direction of Gradient
Neilson Mackay
9nem at qlink.queensu.ca
Thu Jul 10 09:06:09 EDT 2008
It was based on a problem I am having.
I have mad ea modified metric that takes a set of points and multiple
3d images. Each image is associated with a transform that transform
the image somewhere in 3d space.
Basically I have 1 Registration transform and N image transforms ( n
= nmber of images):
Metric::GetValue(transform registration_transform)
{
for each point p in Pointset
{
for each image Im in set of images ImageSet
{
registration_transform.TransformPoint(p);
im->getTransform->TransformPoint(p);
... normal metric stuff
}
}
}
The problem I am hainvg is I am not sure how to calculate the
derivitive. I assume in normal registration, the "jacobian"
transforms the gradient. If I have two transforms should I
concatenate them together then create a jacobain from that? Or
multiply the seprate jacobains together?
Does this make sence? :) Thanks for the help!
Neilson
On Jul 10, 2008, at 5:41 AM, Luis Ibanez wrote:
>
> Hi Neilson,
>
> It "should" be independent.
>
> But this is something for which we don't have a
> Nightly test as verification...
>
>
> Is your question a theoretical one ?
>
> or do you have a specific ITK filter in mind ?
>
>
> Please let us know,
>
>
> Thanks
>
>
> Luis
>
>
> ----------------------
> Neilson Mackay wrote:
>> Hi,
>> Is the rotation component of a Gradient dependent on the origin
>> of the image? Or is completely independent?
>> Neilson
>> _______________________________________________
>> Insight-users mailing list
>> Insight-users at itk.org
>> http://www.itk.org/mailman/listinfo/insight-users
More information about the Insight-users
mailing list