Scaling coordinate lists

Scaling coordinate lists#

Presume you have a list of coordinates derived from an image. In case the image has a specified pixel size, you can convert pixel coordinates in physical coordinates.

from skimage.io import imread
import pyclesperanto_prototype as cle
import stackview

image = imread("../../data/blobs.tif")[:50,:50]

label_image = cle.voronoi_otsu_labeling(image, spot_sigma=4)

label_image

cle._ image

shape	(50, 50)
dtype	uint32
size	9.8 kB
min	0.0
max	6.0

From the objects in this label image, we can derive centroid coordinates.

centroids = cle.centroids_of_labels(label_image)
centroids

cle.array([[ 0.7586207  7.7894735 17.849672  24.59091   29.01266   43.57143  ]
 [26.275862  44.63158   17.79085    4.071429  47.02532   26.266666 ]], dtype=float32)

Such a point list has d times n numbers for d dimensions (commonly 2 or 3) and n points.

centroids.shape

(2, 6)

In case pixel size is known, we can compute the centroid coordinates in physical units.

pixel_size_in_microns = [0.1, 0.1]

centroids_in_microns = centroids * cle.asarray([pixel_size_in_microns]).T

centroids_in_microns

cle.array([[0.07586207 0.77894735 1.7849673  2.459091   2.9012659  4.357143  ]
 [2.6275861  4.463158   1.779085   0.40714288 4.702532   2.6266668 ]], dtype=float32)