|
a |
|
b/ants/math/get_centroids.py |
|
|
1 |
__all__ = ["get_centroids"] |
|
|
2 |
|
|
|
3 |
import numpy as np |
|
|
4 |
import ants |
|
|
5 |
from ants.decorators import image_method |
|
|
6 |
|
|
|
7 |
@image_method |
|
|
8 |
def get_centroids(image, clustparam=0): |
|
|
9 |
""" |
|
|
10 |
Reduces a variate/statistical/network image to a set of centroids |
|
|
11 |
describing the center of each stand-alone non-zero component in the image |
|
|
12 |
|
|
|
13 |
ANTsR function: `getCentroids` |
|
|
14 |
|
|
|
15 |
Arguments |
|
|
16 |
--------- |
|
|
17 |
image : ANTsImage |
|
|
18 |
image from which centroids will be calculated |
|
|
19 |
|
|
|
20 |
clustparam : integer |
|
|
21 |
look at regions greater than or equal to this size |
|
|
22 |
|
|
|
23 |
Returns |
|
|
24 |
------- |
|
|
25 |
ndarray |
|
|
26 |
|
|
|
27 |
Example |
|
|
28 |
------- |
|
|
29 |
>>> import ants |
|
|
30 |
>>> image = ants.image_read( ants.get_ants_data( "r16" ) ) |
|
|
31 |
>>> image = ants.threshold_image( image, 90, 120 ) |
|
|
32 |
>>> image = ants.label_clusters( image, 10 ) |
|
|
33 |
>>> cents = ants.get_centroids( image ) |
|
|
34 |
""" |
|
|
35 |
imagedim = image.dimension |
|
|
36 |
if clustparam > 0: |
|
|
37 |
mypoints = ants.label_clusters(image, clustparam, max_thresh=1e15) |
|
|
38 |
if clustparam == 0: |
|
|
39 |
mypoints = image.clone() |
|
|
40 |
mypoints = ants.label_stats(mypoints, mypoints) |
|
|
41 |
nonzero = mypoints[["LabelValue"]] > 0 |
|
|
42 |
mypoints = mypoints[nonzero["LabelValue"]] |
|
|
43 |
mypoints = mypoints.iloc[:, :] |
|
|
44 |
x = mypoints.x |
|
|
45 |
y = mypoints.y |
|
|
46 |
|
|
|
47 |
if imagedim == 3: |
|
|
48 |
z = mypoints.z |
|
|
49 |
else: |
|
|
50 |
z = np.zeros(mypoints.shape[0]) |
|
|
51 |
|
|
|
52 |
if imagedim == 4: |
|
|
53 |
t = mypoints.t |
|
|
54 |
else: |
|
|
55 |
t = np.zeros(mypoints.shape[0]) |
|
|
56 |
|
|
|
57 |
centroids = np.stack([x, y, z, t]).T |
|
|
58 |
return centroids |