Skip to content

CLI Statismo Build GP Model

Roman Shapovalov edited this page Jul 14, 2015 · 4 revisions

NAME

statismo-build-gp-model - builds shape or deformation models from a given gaussian process definition

SYNOPSIS

statismo-build-gp-model [options] -r reference-file output-file

DESCRIPTION

statismo-build-gp-model is used to build a shape or deformation model from a gaussian process definition.

OPTIONS

Mandatory

-t, --type TYPE : Specify the type of the model. TYPE can either be shape or deformation.

-d, --dimensionality : Specifies the dimensionality of the input image (either 2 or 3). This option is only available if the type if deformation

-k, --kernel KERNEL : Specify the kernel (covariance function) of the gaussian process. Supported kernels: gaussian

-p, --parameters KERNEL_PARAMETERS : KERNEL_PARAMETERS are the kernel parameters. The exact parameters depend on the kernel:

- gaussian: The gaussian kernel has one parameter
	- sigma: This parameter is a float that sets the sigma parameter in the gaussian kernel. Bigger values make the changes "smoother".

-s, --scale SCALE : SCALE is a scaling factor the kernel is scaled with. The bigger this value is, the stronger the change is when modifying a model parameter.

-n, --numberofbasisfunctions NUMBER_OF_BASIS_FUNCTIONS : NUMBER_OF_BASIS_FUNCTIONS specifies how many parameters the model will have.

-r, --reference REFERENCE_FILE : REFERENCE_FILE is the path to the reference file (a mesh or an image) that will be used to construct the model.

-o, --output-file OUTPUT_FILE : OUTPUT_FILE is the path where the model will be saved.

Optional

-m, --input-model MODEL_FILE : MODEL_FILE Extends an existing model with data from the specified kernel. This is useful to extend existing models in case of insufficient data.

Examples

Build a shape model with a gaussian kernel and 50 basis functions:

statismo-build-gp-model -t shape -k gaussian -p 95.5 -s 100.42 -n 50 -r reference-mesh.vtk model.h5

Build a 3D deformation model with a gaussian kernel and 25 basis functions:

statismo-build-gp-model -t deformation -k gaussian -p 75 -s 100 -n 25 -r reference-image.vtk model.h5

Build a 2D deformation model with a gaussian kernel and 25 basis functions:

statismo-build-gp-model -d 2 -t deformation -k gaussian -p 75 -s 100 -n 25 -r reference-image.vtk model.h5

Extend a 2D deformation model with a gaussian kernel and scale it up to 100 basis functions:

statismo-build-gp-model -d 2 -t deformation -k gaussian -p 75 -s 100 -n 100 -r reference-image.vtk -m input-model.h5 extended-model.h5

SEE ALSO

##Building Models: statismo-build-shape-model (8). Builds shape models from a list of meshes.

statismo-build-deformation-model (8). Builds deformation models from a list of deformation fields

statismo-build-gp-model (8). Builds shape or deformation models from a given gaussian process definition.

##Working with models:

statismo-sample (8). Draws samples from a model.

statismo-reduce-model (8). Reduces the number of components in a model.

statismo-posterior (8). Creates a posterior model from an existing model.

statismo-fit-surface (8). Fits a model iteratively in to a target mesh.

statismo-fit-image (8). Fits a model iteratively to an image.

statismo-warp-image (8). Applies a deformation field to an image.