Hi everybody!
I just started to play with 3D GMM but I have some questions. Not sure I am
doing the right things.
I landmarked (anatomical+curves) the left side of six skulls and I would like
to get my final matrix to do my analyses. I defined my landmarks and subsampled
the landmarks on the curves. Then I patched the landmarks on all my specimens.
So, for now, I have my 6 specimens with the landmarks at the right place on
them.
And here begin the troubles… I wanted to relax and slide the landmarks to
minimize bending energy but I think the surfaces are too different because
unless I use a stepsize of 0 (meaning that this command is useless I think) in
relaxing and sliding commands, the landmarks do not fit the curves anymore.
#relaxLM(patched[,,i], rbind(atlas[["landmarks"]],atlas[["patch"]]), slidings,
outlines = curv, surp = surface, sur.name = NULL, mesh =
ply2mesh(paste("./ply/",meshlist[i],sep="")), tol = 1e-50, deselect = FALSE,
inc.check = F, iterations = 50, fixRepro = TRUE, stepsize=0)
#slider3d(relaxedtemplate4, slidings, outlines = curv, surp = surface, sur.path
= "./ply", sur.name = NULL, meshlist = paste("./ply/",meshlist,sep=""), ignore
= NULL, sur.type = "ply", tol = 10e-50, deselect = FALSE, inc.check = FALSE,
recursive = TRUE, iterations = 3, initproc = TRUE, pairedLM = 0, mc.cores = 4,
fixRepro = TRUE, stepsize =0)
I tried to play with the tolerance and/or the bending=TRUE but nothing has
changed.
What could I change in these commands? Are these steps mandatory to analyse the
shape?
Any comment, advice and suggestion is more than welcome!
Thanks a lot for your help.
Aurélien Lowie
--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
---
To unsubscribe from this group and stop receiving emails from it, send an email
to morphmet+unsubscr...@morphometrics.org.
