def test_grain():
L = np.array([
[0,0,1,1,1,0],
[0,1,0,0,0,0],
[0,0,2,2,2,0],
[0,2,2,2,2,2],
])
D = np.array([
[0, 0,12,12,12,0],
[0,12, 0, 0, 0,0],
[0,0,21,22,23,0],
[0,24,25,26,27,28],
])
assert np.all( pymorph.grain(D, L, 'max', 'data') == (12, 28))
assert np.all( pymorph.grain(D, L, 'min', 'data') == (12, 21))
assert np.all( pymorph.grain(D, L, 'mean', 'data') == (12, 24.5))
def test_image():
f = np.array([
[0,1,2,3,0],
[0,1,3,4,1],
])
labels = np.array([
[1,1,2,2,0],
[1,1,2,2,0],
])
for measure in ('max', 'mean', 'min', 'std'):
image = pymorph.grain(f, labels, measure, 'image')
data = pymorph.grain(f, labels, measure, 'data')
assert len(data) == labels.max()
for v,lab in zip(image.ravel(), labels.ravel()):
if lab > 0:
assert v == data[lab-1]
def regionProps(mask, image=None):
"""Calculates some basic properties of ROIs in a mask.
Properties calculated: centroids, boxes, areas. If the original image is passed in,
this also calcluate the mean value in each region.
Keys of returned dictionary: 'centroid', 'boundingBox', 'area', and optionally, 'meanIntensity'
:param mask: a 2d labeled image
:param image: an optional 2p numpy array, original image, for calculation of mean intensity values
:returns: a dictionary of lists, each containing property values
"""
if image is not None:
if len(image.shape) >= 3:
meanImage = np.mean(image, axis=2)
else:
meanImage = image
min = meanImage.min()
if min < 0:
meanImage -= meanImage.min()
means = pymorph.grain(meanImage,
labels=mask,
measurement='mean',
option='data')
else:
means = None
numLabels = mask.max()
centroids = pymorph.blob(mask, measurement='centroid', output='data')
boxes = pymorph.blob(mask, measurement='boundingbox', output='data')
area = pymorph.blob(mask, measurement='area', output='data')
coms = []
for i in range(1, numLabels + 1):
coms.append(nd.measurements.center_of_mass(meanImage,
labels=mask == i))
if means is not None:
props = [{
'meanIntensity': means[i],
'centroid': centroids[i],
'com': coms[i],
'boundingBox': boxes[i],
'area': area[i]
} for i in range(numLabels)]
else:
props = [{
'centroid': centroids[i],
'com': coms[i],
'boundingBox': boxes[i],
'area': area[i]
} for i in range(numLabels)]
return props
def test_off_by_one():
# Contributed by Timothy Hirzel
L = np.array([
[0,0,1,1,1,0],
[0,1,0,0,0,0],
[0,0,2,2,2,0],
[0,2,2,2,2,2],
])
D = np.array([
[0, 0,12,12,12, 0],
[0,12, 0, 0, 0, 0],
[0, 0,21,22,23, 0],
[0,24,25,26,27,28],
])
maxdata = np.array([12,28])
maximage = np.array([
[0, 0,12,12,12, 0],
[0,12, 0, 0, 0, 0],
[0, 0,28,28,28, 0],
[0,28,28,28,28,28],
])
assert((maxdata == pymorph.grain(D,L,'max', 'data')).all())
assert((maximage == pymorph.grain(D,L,'max', 'image')).all())
def regionProps(mask, image=None):
"""Calculates some basic properties of ROIs in a mask.
Properties calculated: centroids, boxes, areas. If the original image is passed in,
this also calcluate the mean value in each region.
Keys of returned dictionary: 'centroid', 'boundingBox', 'area', and optionally, 'meanIntensity'
:param mask: a 2d labeled image
:param image: an optional 2p numpy array, original image, for calculation of mean intensity values
:returns: a dictionary of lists, each containing property values
"""
if image is not None:
if len(image.shape) >= 3:
meanImage = np.mean(image, axis=2)
else:
meanImage = image
min = meanImage.min()
if min < 0:
meanImage -= meanImage.min()
means = pymorph.grain(meanImage, labels=mask, measurement='mean', option='data')
else:
means = None
numLabels = mask.max()
centroids = pymorph.blob(mask, measurement='centroid', output='data')
boxes = pymorph.blob(mask, measurement='boundingbox', output='data')
area = pymorph.blob(mask, measurement='area', output='data')
coms = []
for i in range(1, numLabels+1):
coms.append(nd.measurements.center_of_mass(meanImage, labels=mask==i))
if means is not None:
props = [{'meanIntensity':means[i], 'centroid':centroids[i], 'com':coms[i], 'boundingBox':boxes[i], 'area':area[i]} for i in range(numLabels)]
else:
props = [{'centroid':centroids[i], 'com':coms[i], 'boundingBox':boxes[i], 'area':area[i]} for i in range(numLabels)]
return props
