You received this message because you are subscribed to the Google Groups "scalismo" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scalismo+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/scalismo/4a23e207-56f9-40fd-b279-8e9292695bbcn%40googlegroups.com.
The shape model always deforms the reference mesh. Even when you show the mean of the model, the reference mesh is deformed to depict the mean mesh. So a point x from the reference is shown as z = x + \mu(x) with \mu beiing the mean deformation. If you draw a sample of the model, the mean deformation and the sampled deformation is applied to the reference model. Meaning a point is then z = x + \mu(x) + u(x) if u is the sampled deformation. This deformation \mu+u is applied to all other objects within the same group.
The nice "side" effect of it is: If you click landmarks on the reference and add it to the group, then the landmarks will always end up at the corresponding location of the surface. Also when you use the UI to display different samples from the model.
The unexpected "side" effect of it is: If you add some points of e.g. your mean mesh as landmarks, and you do not have a mean deformation \mu of zero, then these points appear to not lie on the surface. This is because the point on the mean is z = x + \mu(x) and then the UI adds on top of that again the shape model transpormation \mu+u .
So in the shape model group you should only display the reference
mesh or points that are defined on the reference (or on the mean
if your \mu is zero). Everything else you should display in
another group to prevent that the shape model transformation is
applied to it.
I hope this made the mechanism a bit clearer.
To view this discussion on the web visit https://groups.google.com/d/msgid/scalismo/d3140ddc-f333-406f-a668-cee163dc305cn%40googlegroups.com.