pystackreg API reference

class pystackreg.StackReg(transformation)[source]

Python implementation of the ImageJ/Fiji StackReg plugin ( http://bigwww.epfl.ch/thevenaz/stackreg/ )

get_matrix()[source]

Get the current transformation matrix

Return type:ndarray(3,3) or ndarray(4,4) for bilinear transformation
Returns:The transformation matrix
get_points()[source]

Returns the pairs of corresponding points from which the transformation matrix can be calculated.

Return type:(ndarray, ndarray)
Returns:(Point coordinates of reference image, Point coordinates of image to be aligned)
is_registered()[source]

Indicates whether register() was already called and a transformation matrix was calculated

Return type:bool
Returns:True if a transformation matrix was already calculated
register(ref, mov)[source]

Registers an image to a reference image: Only the transformation matrix will be calculated, the image will not be transformed (use transform() or register_transform() ).

Parameters:
  • ref (array_like (Ni..., Nj...)) – Reference image (static)
  • mov (array_like (Ni..., Nj...)) – The image that should be aligned to the reference image
Return type:

ndarray(3,3)
Returns:

2D transformation matrix
register_stack(img, reference='previous', n_frames=1, axis=0, moving_average=1, verbose=False, progress_callback=None)[source]

Register a stack of images (movie). Note that this function will not transform the image but only calculate the transformation matrices. For tranformation use transform_stack() after this function or use register_transform_stack() for a single call.

Parameters:
  • img (array_like(Ni..., Nj..., Nk...)) – The 3D stack of images that should be aligned
  • reference (string, one of ['previous', 'first', 'mean']) –
    • ‘previous’: Aligns each frame (image) to its previous frame in the stack
    • ’first:’ Aligns each frame (image) to the first frame in the stack - if n_frames is > 1, then each frame is aligned to the mean of the first n_frames of the stack
    • ’mean’: Aligns each frame (image) to the average of all images in the stack
  • n_frames (int, optional) – If reference is ‘first’, then this parameter specifies the number of frames from the beginning of the stack that should be averaged to yield the reference image.
  • axis (int, optional) – The axis of the time dimension
  • moving_average (int, optional) – If moving_average is greater than 1, a moving average of the stack is first created (using a subset size of moving_average) before registration
  • verbose (bool, optional) – Specifies whether a progressbar should be shown using tqdm.
  • progress_callback (function, optional) – A function that is called after every iteration. This function should accept the keyword arguments current_iteration:int and end_iteration:int.
Return type:

ndarray(img.shape[axis], 3, 3)
Returns:

The transformation matrix for each image in the stack
register_transform(ref, mov)[source]

Register and transform an image to a reference image.

Parameters:
  • ref (ref: array_like (Ni..., Nj...)) – Reference image (static)
  • mov (array_like (Ni..., Nj...)) – The image that should be aligned to the reference image
Return type:

ndarray (Ni…, Nj…)
Returns:

Transformed image - will be of the same shape as the input image (cropping may occur)
register_transform_stack(img, reference='previous', n_frames=1, axis=0, moving_average=1, verbose=False, progress_callback=None)[source]

Register and transform stack of images (movie).

Parameters:
  • img (array_like(Ni..., Nj..., Nk...)) – The 3D stack of images that should be aligned
  • reference (string, one of ['previous', 'first', 'mean']) –
    • ‘previous’: Aligns each frame (image) to its previous frame in the stack
    • ’first:’ Aligns each frame (image) to the first frame in the stack - if n_frames is > 1, then each frame is aligned to the mean of the first n_frames of the stack
    • ’mean’: Aligns each frame (image) to the average of all images in the stack
  • n_frames (int, optional) – If reference is ‘first’, then this parameter specifies the number of frames from the beginning of the stack that should be averaged to yield the reference image.
  • axis (int, optional) – The axis of the time dimension
  • moving_average (int, optional) – If moving_average is greater than 1, a moving average of the stack is first created (using a subset size of moving_average) before registration
  • verbose (bool, optional) – Specifies whether a progressbar should be shown using tqdm.
  • progress_callback (function, optional) – A function that is called after every iteration. This function should accept the keyword arguments current_iteration:int and end_iteration:int.
Return type:

ndarray(Ni…, Nj…, Nk…)
Returns:

The transformed stack
set_matrix(mat)[source]

Sets the current transformation matrix

Parameters:mat (ndarray(3,3) or ndarray(4,4) for bilinear transformation) – The transformation matrix
transform(mov, tmat=None)[source]

Transform an image according to a previous registration.

Either a transformation matrix has to be explicitly supplied or register() has to be called before calling transform().

Parameters:
  • mov (array_like (Ni..., Nj...)) – The image that will be transformed
  • tmat (ndarray(3,3), optional) – The transformation matrix
Return type:

ndarray (Ni…, Nj…)
Returns:

Transformed image - will be of the same shape as the input image (cropping may occur)
transform_stack(img, axis=0, tmats=None)[source]

Transform a stack after registration.

Parameters:
  • img (array_like(Ni..., Nj..., Nk...)) – The 3D stack of images that should be aligned
  • axis (int, optional) – The axis of the time dimension
  • tmats (ndarray(img.shape[axis], 3, 3), optional) – The transformation matrix for each image in the stack
Return type:

ndarray(Ni…, Nj…, Nk…)
Returns:

The transformed stack

Indices and tables