sksurgerycalibration.algorithms.pivot module

Functions for pivot calibration.

sksurgerycalibration.algorithms.pivot.pivot_calibration(tracking_matrices, configuration=None)[source]

Performs pivot calibration on an array of tracking matrices

Parameters:
  • tracking_matrices – an Nx4x4 array of tracking matrices
  • configuration – an optional configuration dictionary, if not the algorithm defaults to Algebraic One Step. Other options include ransac, and sphere_fitting
Returns:

tuple containing; ‘pointer_offset’ The coordinate of the pointer tip relative to the tracking centre ‘pivot_point’ The location of the pivot point in world coordinates ‘residual_error’ The RMS pointer tip error, errors in each direction are treated as independent variables, so for a calibration with n matrices, RMS error is calculated using nx3 measurements.
Raises:

TypeError, ValueError
sksurgerycalibration.algorithms.pivot.pivot_calibration_aos(tracking_matrices)[source]

Performs Pivot Calibration, using Algebraic One Step method, and returns Residual Error.

See Yaniv 2015.

Parameters:tracking_matrices – N x 4 x 4 ndarray, of tracking matrices.
Returns:pointer offset, pivot point and RMS Error about centroid of pivot.
Raises:ValueError if rank less than 6
sksurgerycalibration.algorithms.pivot.pivot_calibration_sphere_fit(tracking_matrices, init_parameters=None)[source]

Performs Pivot Calibration, using sphere fitting, based on

See Yaniv 2015.

Parameters:
  • tracking_matrices – N x 4 x 4 ndarray, of tracking matrices.
  • init_parameters – 1X4 array of initial parameter for finding the pivot point in world coords and pivot radius. Default is to set to the mean x,y,z values and radius = 0.
Returns:

pointer offset, pivot point and RMS Error about centroid of pivot.
sksurgerycalibration.algorithms.pivot.pivot_calibration_with_ransac(tracking_matrices, number_iterations, error_threshold, concensus_threshold, early_exit=False)[source]

Written as an exercise for implementing RANSAC.

Parameters:
  • tracking_matrices – N x 4 x 4 ndarray, of tracking matrices.
  • number_iterations – the number of iterations to attempt.
  • error_threshold – distance in millimetres from pointer position
  • concensus_threshold – the minimum percentage of inliers to finish
  • early_exit – If True, returns model as soon as thresholds are met
Returns:

pointer offset, pivot point and RMS Error about centroid of pivot.
Raises:

TypeError, ValueError