loki.align.DeformableRegistration

class loki.align.DeformableRegistration(alpha=None, beta=None, low_rank=False, num_eig=100, *args, **kwargs)

Deformable registration. Adapted from Pure Numpy Implementation of the Coherent Point Drift Algorithm: https://github.com/siavashk/pycpd

alpha

Represents the trade-off between the goodness of maximum likelihood fit and regularization.

Type:

float (positive)

beta

Width of the Gaussian kernel.

Type:

float(positive)

low_rank

Whether to use low rank approximation.

Type:

bool

num_eig

Number of eigenvectors to use in lowrank calculation.

Type:

int

__init__(alpha=None, beta=None, low_rank=False, num_eig=100, *args, **kwargs)

Methods

__init__([alpha, beta, low_rank, num_eig])
expectation()	Compute the expectation step of the EM algorithm.
get_registration_parameters()	Return the current estimate of the deformable transformation parameters.
iterate()	Perform one iteration of the EM algorithm.
maximization()	Compute the maximization step of the EM algorithm.
register([callback])	Perform the EM registration.
transform_point_cloud([Y])	Update a point cloud using the new estimate of the deformable transformation.
update_transform()	Calculate a new estimate of the deformable transformation.
update_variance()	Update the variance of the mixture model using the new estimate of the deformable transformation.