prealignment¶

tiatoolbox . tools . registration . wsi_registration . prealignment

prealignment(fixed_img, moving_img, fixed_mask, moving_mask, dice_overlap=0.5, rotation_step=10)[source]¶

Coarse registration of an image pair.

This function performs initial alignment of a moving image with respect to a fixed image. This can be used as a prealignment step before final refinement.

Parameters:

fixed_img (numpy.ndarray) – A fixed image.
moving_img (numpy.ndarray) – A moving image.
fixed_mask (numpy.ndarray) – A binary tissue mask for the fixed image.
moving_mask (numpy.ndarray) – A binary tissue mask for the moving image.
dice_overlap (float) – Dice ratio used for the selection of the best transformation matrix.
rotation_step (int) – Rotation_step defines an increment in the rotation angles.

Returns:

numpy.ndarray - A rigid transform matrix.
numpy.ndarray - Transformed moving image.
numpy.ndarray - Transformed moving mask.
float - Dice overlap.

Return type:

tuple

Examples

>>> from tiatoolbox.tools.registration.wsi_registration import prealignment
>>> transform, transformed_image, transformed_mask, dice_overlap = prealignment(
...     fixed_thumbnail, moving_thumbnail, fixed_mask, moving_mask
... )