pystack3d.registration_calculation module

Functions related to the registration calculation processing

pystack3d.registration_calculation.init_args(params, shape)

Initialize arguments related to the current processing (‘registration_calculation’) and return a specific array to share (tmats)

Parameters:

  • params (dict) – Dictionary related to the current process. See the related documentation for more details.

  • shape (tuple of 3 int) – Shape of the stack to process

Returns:

tmats – Transformation matrices to be shared during the (multi)processing

Return type:

numpy.ndarray((shape[0], nblocks[0]*nblocks[1], 3, 3))

pystack3d.registration_calculation.registration_calculation(fnames=None, inds_partition=None, queue_incr=None, area=None, threshold=None, nb_blocks=None, transformation='TRANSLATION', output_dirname=None)

Function dedicated to the transformation matrices calculation used in the image registration

Parameters:

  • fnames (list of pathlib.Path, optional) – List of ‘.tif’ filenames to process

  • inds_partition (list of ints, optional) – List of indexes to be considered by the global var SHARED_ARRAY when working in multiprocessing

  • queue_incr (multiprocessing.Queue, optional) – Queue passed to the function to interact with the progress bar

  • area (iterable of 4 ints, optional) – Area defined from coordinates (xmin, xmax, ymin, ymax) to calculate the transformation matrices. If None, consider the whole images.

  • threshold (float, optional) – Threshold value used to perform registration on binarized images

  • nb_blocks (tuple of 2 ints (p, q), optional) – Number of patches (sub-images) to consider in each axis to perform the registrations. If None, consider the whole image

  • transformation (str, optional) – Transformation matrix type to consider for the registration. To choose among : [‘TRANSLATION’, ‘RIGID_BODY’, ‘SCALED_ROTATION’, ‘AFFINE’] (See pystackreg documentation for more details). If None, the default Transformation matrix type is ‘TRANSLATION’

  • output_dirname (str, optional) – Directory pathname for process results saving

pystack3d.registration_calculation.tmat_calculation(ref, img, nb_blocks=None, transformation='TRANSLATION')

Transformation matrix calculation with StackReg from pystackreg package applied on patches (sub-images)

Parameters:

  • ref (numpy.ndarray((m,n))) – Image used as reference

  • img (numpy.ndarray((m,n))) – Image considered as ‘moving’ image

  • nb_blocks (tuple of 2 ints (p, q), optional) – Number of patches (sub-images) to consider in each axis to perform the registrations. If None, consider the whole image

  • transformation (str, optional) – Transformation matrix type to consider for the registration. To choose among : [‘TRANSLATION’, ‘RIGID_BODY’, ‘SCALED_ROTATION’, ‘AFFINE’] (See pystackreg documentation for more details). If None, the default Transformation matrix type is ‘TRANSLATION’

Returns:

tmats – Transformation matrix calculated in all patches

Return type:

numpy.ndarray((p*q, 3*3))

pystack3d.registration_calculation.plot(output_dirname)

Plot the specific data related to the current process

pystack3d.registration_calculation.registration_plot(tmats, nb_blocks=None, transformation='RIGID_BODY', cumul=False, title='Registration')

Plot the evolution of the transformation matrix components issued from the registration processing

Parameters:

  • tmats (sequence of nslices * nb_blocks * numpy.ndarray((3, 3))) – Transformations matrices calculated during the registration processing

  • nb_blocks (tuple of 2 ints (p, q)) – Number of patches (sub-images) used during the registration processing

  • transformation (str, optional) – Transformation matrix type considered for the plotting. If None, consider a ‘RIGID_BODY’ transformation type.

  • cumul (bool, optional) – Activation key to plot the cumulative transformations matrices evolution

  • title (str, optional) – Title of the figure

Returns:

fig

Return type:

matplotlib.figure

pystack3d.registration_calculation.decomposition(tmat)

Return ‘transl_x’, ‘transl_y’, ‘rotation’, ‘scale_x’, ‘scale_y’ and ‘shear’ related to a 2D-Affine transformation matrix

Notes

Translation values are given according to “natural” (x, y) coordinates. Rotation angle are given according to the counter-clockwise convention.