spam.filters package

Submodules

spam.filters.distanceField module

spam.filters.distanceField.distanceField(phases, phaseID=1)

This function tranforms an array/image of integers into a continuous field. It works for segmented binary/trinary 3D images or arrays of integers. It has to be run for each phase seperately.

It uses of the Distance Transform Algorithm. For every voxel belonging to a phase a value indicating the distance (in voxels) of that point to the nearest background point is computed. The DTA is computed for the inverted image as well and the computed distances are setting to negative values. The 2 distance fields are merged into the final continuuos distance field where:

- positive numbers: distances from the phase to the nearest background
voxel
- negative values: distances from the background to the nearest phase
voxel
- zero values: the interface between the considered phase and the
background

Parameters:

• phases (array) – The input image/array (each phase should be represented with only one number)

• phaseID (int, default=1) – The integer indicating the phase which distance field you want to calculate

Returns:

distance field of the phase

Return type:

array

Example

>>> import tifffile
>>> import spam.filters
>>> im = tifffile.imread( "mySegmentedImage.tif" )
In this image the inclusions are labelled 1 and the matrix 0
>>> di = spam.filters.distanceField( im, phase=1 )
The resulting distance field is made of float between -1 and 1

spam.filters.morphologicalOperations module

spam.filters.morphologicalOperations.greyDilation(im, nBytes=1)

This function applies a dilation on a grey scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• nBytes (int, default=1) – Number of bytes used to substitute the values on the border.

Returns:

The dilated image

Return type:

numpy array

spam.filters.morphologicalOperations.greyErosion(im, nBytes=1)

This function applies a erosion on a grey scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• nBytes (int, default=1) – Number of bytes used to substitute the values on the border.

Returns:

The eroded image

Return type:

numpy array

spam.filters.morphologicalOperations.greyMorphologicalGradient(im, nBytes=1)

This function applies a morphological gradient on a grey scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• nBytes (int, default=1) – Number of bytes used to substitute the values on the border.

Returns:

The morphologycal gradient of the image

Return type:

numpy array

spam.filters.morphologicalOperations.binaryDilation(im, sub=False)

This function applies a dilation on a binary scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• sub (bool, default=False) – Subtitute value.

Returns:

The dilated image

Return type:

numpy array

spam.filters.morphologicalOperations.binaryErosion(im, sub=False)

This function applies a erosion on a binary scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• sub (bool, default=False) – Substitute value.

Returns:

The eroded image

Return type:

numpy array

spam.filters.morphologicalOperations.binaryMorphologicalGradient(im, sub=False)

This function applies a morphological gradient on a binary scale image

Parameters:

• im (numpy array) – The input image (greyscale)

• nBytes (int, default=False) – Number of bytes used to substitute the values on the border.

Returns:

The morphologycal gradient of the image

Return type:

numpy array

spam.filters.morphologicalOperations.binaryGeodesicReconstruction(im, marker, dmax=None, verbose=False)

Calculate the geodesic reconstruction of a binary image with a given marker

Parameters:

• im (numpy.array) – The input binary image

• marker (numpy.array or list) – If numpy array: direct input of the marker (must be the size of im) If list: description of the plans of the image considered as the marker | [1, 0] plan defined by all voxels at x1=0 | [0, -1] plan defined by all voxels at x0=x0_max | [0, 0, 2, 5] plans defined by all voxels at x0=0 and x2=5

• dmax (int, default=None) – The maximum geodesic distance. If None, the reconstruction is complete.

• verbose (bool, default=False) – Verbose mode

Returns:

The reconstructed image

Return type:

numpy.array

spam.filters.morphologicalOperations.morphologicalReconstruction(im, selem=array([[[0, 0, 0], [0, 1, 0], [0, 0, 0]], [[0, 1, 0], [1, 1, 1], [0, 1, 0]], [[0, 0, 0], [0, 1, 0], [0, 0, 0]]], dtype=uint8))

This functions performs a morphological reconstruction (greyscale opening followed by greyscale closing). The ouput image presents less variability in the greyvalues inside each phase, without modifying the original shape of the objects of the image. -

Parameters:

• im (3D numpy array) – Greyscale image to perform the reconstuction

• selem (structuring element, optional) – Structuring element Default = None

Returns:

imReconstructed – Greyscale image after the reconstuction

Return type:

3D boolean array

Module contents

spam.filters.distanceField(phases, phaseID=1)

This function tranforms an array/image of integers into a continuous field. It works for segmented binary/trinary 3D images or arrays of integers. It has to be run for each phase seperately.

It uses of the Distance Transform Algorithm. For every voxel belonging to a phase a value indicating the distance (in voxels) of that point to the nearest background point is computed. The DTA is computed for the inverted image as well and the computed distances are setting to negative values. The 2 distance fields are merged into the final continuuos distance field where:

- positive numbers: distances from the phase to the nearest background
voxel
- negative values: distances from the background to the nearest phase
voxel
- zero values: the interface between the considered phase and the
background

Parameters:

• phases (array) – The input image/array (each phase should be represented with only one number)

• phaseID (int, default=1) – The integer indicating the phase which distance field you want to calculate

Returns:

distance field of the phase

Return type:

array

Example

>>> import tifffile
>>> import spam.filters
>>> im = tifffile.imread( "mySegmentedImage.tif" )
In this image the inclusions are labelled 1 and the matrix 0
>>> di = spam.filters.distanceField( im, phase=1 )
The resulting distance field is made of float between -1 and 1