Slicer

Module for slicing two-dimensional arrays to three-dimensional arrays of stacked masks.

Functions:

Name	Description
calculate_region_properties	Calculate the region properties on a segmented array.
mask_slices	Convert a three-dimensional sliced array into masks based.
mask_small_artefacts	Mask labelled features that are less than a specified size from two-dimensional array.
mask_small_artefacts_all_layers	Mask labelled features that are less than a specified size from three-dimensional array.
region_properties_by_slices	Calculate region properties for each layer in a three-dimensinoal sliced array.
segment	Segment an array.
segment_slices	Segment individual layers of a three-dimensional numpy array.
show_layers	Helper function for debugging which shows individual layers of a three-dimensional numpy array.
slice_3d	Convert a two-dimensional array of heights into a three-dimensional binary array.
slicer	Convert a two-dimensional array to a three-dimensional stacked array with copies of the original in each layer.

calculate_region_properties

calculate_region_properties(array, spacing)

Calculate the region properties on a segmented array.

The arrays can be either individual slices from a three-dimensional image or the full three-dimensional array itself. If the later then the resulting attribute area will be a "volume". By including the spacing argument, which should be the pixel_to_nm_scaling attribute of the AFMSlicer object the area/volume is in the actual units measured rather than pixels.

Parameters:

Name	Type	Description	Default
array	NDArray	Array of labelled regions.	required
spacing	float	Pixel to nm scaling.	required

Returns:

Type	Description
list[RegionProperties]	A list of RegionProperties.

mask_slices

mask_slices(
    stacked_array,
    slices=None,
    layers=None,
    min_height=None,
    max_height=None,
)

Convert a three-dimensional sliced array into masks based.

A three-dimensional array is converted to masked layers where each layer indicates whether a position in the two-dimensional cross-section is above the threshold for that layer. Thresholds are determined from the data itself if not explicitly provided.

Parameters:

Name	Type	Description	Default
stacked_array	NDArray	Three-dimensional numpy array of image heights, each layer should be a copy of the original.	required
slices	int	Number of slices to mask. Determined directly from data if not provided (i.e. depth of three-dimensional array).	None
layers	NDArray	Array of height thresholds for each slice. Determined directly from data if not provided.	None
min_height	float64	Minimum height. Determined directly from data if not provided.	None
max_height	float64	Maximum height. Determined directly from data if not provided.	None

Returns:

Type	Description
NDArray[bool]	Three-dimensional array of masks.

mask_small_artefacts

mask_small_artefacts(
    labelled_array, properties, minimum_size
)

Mask labelled features that are less than a specified size from two-dimensional array.

Parameters:

Name	Type	Description	Default
labelled_array	NDArray[int32]	A three-dimensional array with labelled regions.	required
properties	dict[int, Any]	The properties of labelled objects for each layer.	required
minimum_size	float	Minimum size below which labelled regions are removed.	required

Returns:

Type	Description
NDArray[int32]	Masked array with labelled regions smaller than minimum_size masked.

mask_small_artefacts_all_layers

mask_small_artefacts_all_layers(
    labelled_array, properties, minimum_size
)

Mask labelled features that are less than a specified size from three-dimensional array.

Parameters:

Name	Type	Description	Default
labelled_array	NDArray[int32]	A three-dimensional array with labelled regions.	required
properties	dict[int, Any]	The properties of labelled objects for each layer.	required
minimum_size	float	Minimum size below which labelled regions are removed.	required

Returns:

Type	Description
NDArray[int32]	Masked array with labelled regions smaller than minimum_size masked.

region_properties_by_slices

region_properties_by_slices(array, spacing)

Calculate region properties for each layer in a three-dimensinoal sliced array.

Parameters:

Name	Type	Description	Default
array	NDArray[int32]	Three-dimensional sliced and labelled array.	required
spacing	float	Pixel to nanometer scaling.	required

Returns:

Type	Description
dict[int, Any]	Dictionary of regionprops calculated using skimage.

segment

segment(array, method='label', tidy_border=True, **kwargs)

Segment an array.

Parameters:

Name	Type	Description	Default
array	NDArray	Two-dimensional numpy array to segment.	required
method	str	Segmentation method, supports the label (default) and watershed methods implemented by Scikit Image.	'label'
tidy_border	bool	Whether to remove objects that straddle the border of the image.	True
**kwargs	dict[str, Any]	Additional arguments to pass for segmentation.	{}

Returns:

Type	Description
NDArray	Labelled array of objects.

segment_slices

segment_slices(array, method='label', tidy_border=False)

Segment individual layers of a three-dimensional numpy array.

Parameters:

Name	Type	Description	Default
array	NDArray[bool]	Three-dimensional boolean array to be segmented.	required
method	str	Sgementation method to use. Currne options are label (default) and watershed.	'label'
tidy_border	bool	Whether to tidy the border.	False

Returns:

Type	Description
NDArray[bool]	Three-dimensional array of labelled layers.

show_layers

show_layers(array)

Helper function for debugging which shows individual layers of a three-dimensional numpy array.

Parameters:

Name	Type	Description	Default
array	NDArray	Three-dimensional numpy array.	required

slice_3d

slice_3d(array, scaling, binary=False)

Convert a two-dimensional array of heights into a three-dimensional binary array.

Rather than taking a fixed number of slices as with slicer() we here use the observed range and the scaling, which typically comes from the scan itself to determine the appropriate number of slices to take so that they are correctly scaled to the same as the x/y axis of the array.

Parameters:

Name	Type	Description	Default
array	NDArray[float64]	A two-dimensional array of heights.	required
scaling	float	Pixel to nanometre scaling.	required
binary	bool	Whether to create a boolean array. If True each pixel is dichotomised to 0/1 if the height is within the range for that layer. If False the value for that layer is returned, this is useful if you want to use colour gradients when viewing three dimensional plots.	False

Returns:

Type	Description
NDArray[int8, float64]	A three-dimensional array scaled by scaling.

slicer

slicer(heights, slices)

Convert a two-dimensional array to a three-dimensional stacked array with copies of the original in each layer.

Parameters:

Name	Type	Description	Default
heights	NDArray[float64]	Two-dimensional numpy array of heights.	required
slices	int	Number of slices to make.	required

Returns:

Type	Description
NDArray[float64]	Expanded numpy array with the original two-dimensional array copied slices in the third dimension.

Examples:

>>> import numpy as np
>>> simple = np.asarray([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
>>> slicer(heights=simple, slices=2)
array([[[1, 1],
    [2, 2],
    [3, 3]],

[[4, 4], [5, 5], [6, 6]],

[[7, 7], [8, 8], [9, 9]]])