Merging labels according to edge-to-edge-distances

Merging labels according to edge-to-edge-distances#

In this notebook we will merge labels in a label image according to their edge-to-edge distances to each other. Labels close-by will be merged.

See also

import pyclesperanto_prototype as cle
from skimage.io import imread
import numpy as np

For demonstration purposes, we use a modified version of the labels derived from the blobs example-image. We artificially introduce gaps between them.

image = imread("../../data/blobs.tif")
image[:, 80:150] = 0
image[80:130, 100:] = 0

image = cle.asarray(image)
labels = cle.voronoi_otsu_labeling(image, spot_sigma=4, outline_sigma=3)
labels

cle._ image

shape	(254, 256)
dtype	uint32
size	254.0 kB
min	0.0
max	47.0

First, we dilate the labels by half of the maximum distance the edges are allowed to have.

maximum_distance = 12

dilated_labels = cle.dilate_labels(labels, radius=maximum_distance/2)
dilated_labels

cle._ image

shape	(254, 256)
dtype	uint32
size	254.0 kB
min	0.0
max	47.0

We then merge the labels if the touch.

merged_dilated_labels = cle.merge_touching_labels(dilated_labels)
merged_dilated_labels

cle._ image

shape	(254, 256)
dtype	uint32
size	254.0 kB
min	0.0
max	5.0

Afterwards, we mask the merged labels with the original label’s shape. We also convert the result of this operation to 32-bit integer, so that the visualization as label-image works.

merged_labels = (merged_dilated_labels * (labels > 0)).astype(np.uint32)
merged_labels

cle._ image

shape	(254, 256)
dtype	uint32
size	254.0 kB
min	0.0
max	5.0