skimage2.data.binary_blobs#

skimage2.data.binary_blobs(shape, blob_size, *, volume_fraction=0.5, rng=None, boundary_mode='wrap')[source]#

Generate synthetic binary image containing blob-like objects.

Parameters:
shapetuple of (int, â€Ķ)

Shape of the output image.

blob_sizefloat

Typical linear size (diameter) of blobs in pixels. Values smaller than 1 or larger than any length in shape may lead to unexpected results. The actual size is also affected by volume_fraction.

volume_fractionfloat, in interval [0, 1], optional

Fraction of image pixels covered by the blobs. Blobs shrink with smaller fractions, and grow and merge together with larger fractions.

rngint or numpy.random.Generator, optional

Pseudo-random number generator. By default, a PCG64 generator is used (see numpy.random.default_rng()). If rng is an int, it is used to seed the generator.

boundary_mode{‘wrap’, ‘nearest’}, optional

The blobs are created by smoothing and then thresholding an array consisting of ones at seed positions. This mode determines which values are filled in when the smoothing kernel overlaps the seed array’s boundary.

‘wrap’ (a b c d | a b c d | a b c d)

By default, the seed array is extended by wrapping around to the opposite edge. The resulting blob array can be tiled and blobs will be contiguous and have smooth edges across tile boundaries.

‘nearest’ (a a a a | a b c d | d d d d)

When applying the Gaussian filter, the seed array is extended by replicating the last boundary value. This will increase the size of blobs whose seed or center lies exactly on the edge.

Returns:
blobsndarray of dtype bool

Output binary image.

Examples

>>> import _skimage2 as ski2
>>> ski2.data.binary_blobs(shape=(5, 5), blob_size=1)
array([[ True, False,  True,  True,  True],
       [ True,  True,  True, False,  True],
       [False,  True, False,  True,  True],
       [ True, False, False,  True,  True],
       [ True, False, False, False,  True]])
>>> blobs = ski2.data.binary_blobs(shape=(256, 256), blob_size=25)
>>> # Finer structures
>>> blobs = ski2.data.binary_blobs(shape=(256, 256), blob_size=13)
>>> # Blobs cover a smaller volume fraction of the image
>>> blobs = ski2.data.binary_blobs(
...     shape=(256, 256), blob_size=25, volume_fraction=0.3
... )