[Insight-users] Registration Framework

[Lisa Dean]

Hi Luis,


Yes, I suppose that would work. But, my input images happen to be in the
form of pointsets/meshes, so I'd require a PointsetToPointset metric (in
the case that I don't want to convert the mesh into an image) right?

And another thing I'd like to be clear on is how the cost function is
passed to the optimizer. I believe the metric does that function but where
does the metric get the cost function from? I mean the metric just
calculates a similarity measure right? Whereas, the cost function might be
an energy function that has to be minimized.
for eg:- the levenbergmarquardt optimizer has a member function called
setcostfunction(MultipleValuedCostFunction* costFunction), but it doesn't
seem to have been used in the IterativeClosestPoint2.cxx. In the case that
this member function is used what (type) would the argument have to be?

Thanks for your help
Lisa
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20120213/68335aec/attachment.htm>