RvtkStatismo: Add isotropic scaling when creating a Gaussian Process model
27 Oct 2014I wondered whether it might be useful to add isotropic scaling to when building a Gaussian Process model from a reference. Because in samples with a fair amount of isotropic variation, the additional information might be worth accounting for.
In RvtkStatismo’s functions statismoModelFromRepresenter and statismoGPmodel, I added the option isoScale to specify the range of the isotropic scale (e.g. isoScale=0.05 will result in a scale factor of ±15% of the original size within the range of 3 standard deviations).
Here is an example (once again using Marcel’s example femurs) - you will, of course need to install latest master branch of RvtkStatismo:
require(RvtkStatismo)
###get the data
download.file(url="https://github.com/marcelluethi/statismo-shaperegistration/raw/master/data/VSD001_femur.vtk","./VSD001_femur.vtk",method = "w")
download.file(url="https://github.com/marcelluethi/statismo-shaperegistration/raw/master/data/VSD002_femur.vtk","./VSD002_femur.vtk",method = "w")
download.file(url="https://github.com/marcelluethi/statismo-shaperegistration/raw/master/data/VSD001-lm.csv","./VSD001-lm.csv",method = "w")
download.file(url="https://github.com/marcelluethi/statismo-shaperegistration/raw/master/data/VSD002-lm.csv","./VSD002-lm.csv",method = "w")
### read the data
ref <- read.vtk("VSD001_femur.vtk")
tar <- read.vtk("VSD002_femur.vtk")
ref.lm <- as.matrix(read.csv("VSD001-lm.csv",row.names=1))
tar.lm <- as.matrix(read.csv("VSD002-lm.csv",row.names=1))
### create models
mymod <- statismoModelFromRepresenter(ref,kernel=list(c(50,50)),ncomp = 100)#default Gaussian Process model
mymodScale <- statismoModelFromRepresenter(ref,kernel=list(c(50,50)),isoScale=0.05,ncomp = 100)#add isotropic scaling with 3 sd being +- 15%
## calculate constrained models based on target landmarks
cmod <- statismoConstrainModel(mymod,tar.lm,ref.lm,2)
cmodS <- statismoConstrainModel(mymodScale,tar.lm,ref.lm,2)
Here are the results (green=reference/constrained model mean):
