def _region_features_for(histone, dna, region):
pixels0 = histone[region].ravel()
pixels1 = dna[region].ravel()
bin0 = pixels0 > histone.mean()
bin1 = pixels1 > dna.mean()
overlap = [np.corrcoef(pixels0, pixels1)[0, 1], (bin0 & bin1).mean(), (bin0 | bin1).mean()]
spi = mh.sobel(histone, just_filter=1)
sp = spi[mh.erode(region)]
sdi = mh.sobel(dna, just_filter=1)
sd = sdi[mh.erode(region)]
sobels = [
np.dot(sp, sp) / len(sp),
np.abs(sp).mean(),
np.dot(sd, sd) / len(sd),
np.abs(sd).mean(),
np.corrcoef(sp, sd)[0, 1],
np.corrcoef(sp, sd)[0, 1] ** 2,
sp.std(),
sd.std(),
]
return np.concatenate(
[
[region.sum()],
haralick(histone * region, ignore_zeros=True).mean(0),
haralick(dna * region, ignore_zeros=True).mean(0),
overlap,
sobels,
haralick(mh.stretch(sdi * region), ignore_zeros=True).mean(0),
haralick(mh.stretch(spi * region), ignore_zeros=True).mean(0),
]
)
def _region_features_for(histone, dna, region):
pixels0 = histone[region].ravel()
pixels1 = dna[region].ravel()
bin0 = pixels0 > histone.mean()
bin1 = pixels1 > dna.mean()
overlap = [
np.corrcoef(pixels0, pixels1)[0, 1],
(bin0 & bin1).mean(),
(bin0 | bin1).mean(),
]
spi = mh.sobel(histone, just_filter=1)
sp = spi[mh.erode(region)]
sdi = mh.sobel(dna, just_filter=1)
sd = sdi[mh.erode(region)]
sobels = [
np.dot(sp, sp) / len(sp),
np.abs(sp).mean(),
np.dot(sd, sd) / len(sd),
np.abs(sd).mean(),
np.corrcoef(sp, sd)[0, 1],
np.corrcoef(sp, sd)[0, 1]**2,
sp.std(),
sd.std(),
]
return np.concatenate([
[region.sum()],
haralick(histone * region, ignore_zeros=True).mean(0),
haralick(dna * region, ignore_zeros=True).mean(0),
overlap,
sobels,
haralick(mh.stretch(sdi * region), ignore_zeros=True).mean(0),
haralick(mh.stretch(spi * region), ignore_zeros=True).mean(0),
])
def test_stretch_rgb():
r = np.arange(256).reshape((32,-1))
g = 255-r
b = r/2
s = mh.stretch(np.dstack([r,g,b]))
s_rgb = mh.stretch_rgb(np.dstack([r,g,b]))
assert not np.all(s == s_rgb)
assert np.all(s[:,:,0] == s_rgb[:,:,0])
assert np.all(mh.stretch(b) == mh.stretch_rgb(b))
def test_stretch_rgb():
r = np.arange(256).reshape((32, -1))
g = 255 - r
b = r / 2
s = mh.stretch(np.dstack([r, g, b]))
s_rgb = mh.stretch_rgb(np.dstack([r, g, b]))
assert not np.all(s == s_rgb)
assert np.all(s[:, :, 0] == s_rgb[:, :, 0])
assert np.all(mh.stretch(b) == mh.stretch_rgb(b))
def calc_img_haralick(im, bins=32):
"""
Calculates haralick features for RGB input image
Linearly quantizes input image to 'bins'.
Averages haralick features across 4 directions.
:param im:
:return: list of RGB haralick features
"""
r_feats = mh.features.haralick(mh.stretch(im[:,:,0], 0, bins), return_mean=True)
g_feats = mh.features.haralick(mh.stretch(im[:,:,1], 0, bins), return_mean=True)
b_feats = mh.features.haralick(mh.stretch(im[:,:,2], 0, bins), return_mean=True)
return r_feats.tolist() + g_feats.tolist() + b_feats.tolist()
def method2(image, sigma):
image = mh.imread(image)[:, :, 0]
image = mh.gaussian_filter(image, sigma)
image = mh.stretch(image)
binimage = image > mh.otsu(image)
labeled, _ = mh.label(binimage)
return labeled
def method2(image, sigma):
image = mh.imread(image)[:, :, 0]
image = mh.gaussian_filter(image, sigma)
image = mh.stretch(image)
binimage = (image > mh.otsu(image))
labeled, _ = mh.label(binimage)
return labeled
def test_run_distance(image, module, image_set, workspace):
module.use_advanced.value = False
module.operation.value = "Distance"
module.x_name.value = "binary"
module.y_name.value = "watershed"
module.footprint.value = 3
data = image.pixel_data
if image.multichannel:
data = skimage.color.rgb2gray(data)
threshold = skimage.filters.threshold_otsu(data)
binary = data > threshold
image_set.add(
"binary",
cellprofiler_core.image.Image(
image=binary, convert=False, dimensions=image.dimensions
),
)
module.run(workspace)
original_shape = binary.shape
distance = scipy.ndimage.distance_transform_edt(binary)
distance = mahotas.stretch(distance)
surface = distance.max() - distance
if image.volumetric:
footprint = numpy.ones((3, 3, 3))
else:
footprint = numpy.ones((3, 3))
peaks = mahotas.regmax(distance, footprint)
if image.volumetric:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16, 16)))
else:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16)))
expected = mahotas.cwatershed(surface, markers)
expected = expected * binary
expected = skimage.measure.label(expected)
actual = workspace.get_objects("watershed")
actual = actual.segmented
numpy.testing.assert_array_equal(expected, actual)
def perform_watershed(threshed, maxima):
distances = mh.stretch(mh.distance(threshed))
spots, n_spots = mh.label(maxima, Bc=np.ones((3, 3)))
surface = (distances.max() - distances)
return sk.morphology.watershed(surface, spots, mask=threshed)
def test_run_distance(image, module, image_set, workspace):
module.operation.value = "Distance"
module.x_name.value = "binary"
module.y_name.value = "watershed"
module.connectivity.value = 3
data = image.pixel_data
if image.multichannel:
data = skimage.color.rgb2gray(data)
threshold = skimage.filters.threshold_otsu(data)
binary = data > threshold
image_set.add(
"binary",
cellprofiler.image.Image(
image=binary,
convert=False,
dimensions=image.dimensions
)
)
module.run(workspace)
original_shape = binary.shape
distance = scipy.ndimage.distance_transform_edt(binary)
distance = mahotas.stretch(distance)
surface = distance.max() - distance
if image.volumetric:
footprint = numpy.ones((3, 3, 3))
else:
footprint = numpy.ones((3, 3))
peaks = mahotas.regmax(distance, footprint)
if image.volumetric:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16, 16)))
else:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16)))
expected = mahotas.cwatershed(surface, markers)
expected = expected * binary
expected = skimage.measure.label(expected)
actual = workspace.get_objects("watershed")
actual = actual.segmented
numpy.testing.assert_array_equal(expected, actual)
def watershed(self, Ta=0):
"""
Identification of particles through inverted slope comparisons
Parameters
-----------
Ta : int
Threshold value in which the particles will be identified by
"""
self.Ta = Ta
dist = mh.distance(self.image > 0.05 * self.Ta)
dist1 = dist
dist = dist.max() - dist
dist -= dist.min() # inverting color
dist = dist / float(dist.ptp()) * 255
dist = dist.astype(np.uint8)
self.dist = mh.stretch(dist, 0, 255)
self.labels, self.n_particles = mh.label(self.image > 0.7 * self.Ta)
thr = np.median(dist)
# not accurate to particles detected(?)
# but matches dist graph well
thresh = (dist < thr)
areas = 0.9 * mh.cwatershed(dist, self.labels)
self.areas = areas * thresh
return
def get_Haralick(im_arr, dist=1, grid_size=1, j=10):
"""
Haralick para una imagen.
:param im_arr:
:param dist:
:param grid_size:
:return: 13*4*grid_size^2 array length
"""
# if j % 30 == 0:
# print("|", end = "", flush = True)
img = np.asarray(im_arr).astype(int)
img = mahotas.stretch(img, 31)
window_size = (np.asarray([img.shape]) / grid_size).astype(int)[0]
im_grid = np.asarray(skimage.util.view_as_blocks(img, tuple(window_size)))
windows = []
for i in range(grid_size):
for j in range(grid_size):
windows.append(im_grid[i, j])
haralick_features = []
for i in range(len(windows)):
try:
h = haralick(windows[i], distance=dist)
except ValueError:
raise
print('error in Haralick!')
print(windows[i].shape)
print(windows[i])
quit()
h = np.ravel(np.asarray(h))
haralick_features.append(h)
out = np.ravel(np.asarray(haralick_features))
return out
def create_ring(self):
im = self._image
# This breaks up the image into RGB channels
r, g, b = im.transpose(2, 0, 1)
h, w = r.shape
# smooth the image per channel:
r12 = mh.gaussian_filter(r, 12.)
g12 = mh.gaussian_filter(g, 12.)
b12 = mh.gaussian_filter(b, 12.)
# build back the RGB image
im12 = mh.as_rgb(r12, g12, b12)
X, Y = np.mgrid[:h, :w]
X = X - h / 2.
Y = Y - w / 2.
X /= X.max()
Y /= Y.max()
# Array C will have the highest values in the center, fading out to the edges:
C = np.exp(-2. * (X ** 2 + Y ** 2))
C -= C.min()
C /= C.ptp()
C = C[:, :, None]
# The final result is sharp in the centre and smooths out to the borders:
ring = mh.stretch(im * C + (1 - C) * im12)
return ring
def ChooseFile(self):
options = QFileDialog.Options()
options |= QFileDialog.DontUseNativeDialog
fileName, _ = QFileDialog.getOpenFileName(self, "QFileDialog.getOpenFileName()", "",
"Excel Files (*.xls *.xlsx *.xlsm);; "
"Tiff Image Files (*.tif);;"
"Other Image Files (*.jpg *png);;"
" All Files (*)", options=options)
if fileName:
if fileName.endswith('.xls') or fileName.endswith('.xlsx') or fileName.endswith('.xlsm'):
self.Line_filename.setText('File Uploaded from: '+str(fileName))
self.seq_dataset = pd.read_excel(fileName, index_col=0);
self.Button_Run.setEnabled(True); self.Button_ShowImage.setEnabled(False);
self.Slider_Sigma.setEnabled(False); self.Slider_Dilate.setEnabled(False);
elif fileName.endswith('.tif') or fileName.endswith('.png') or fileName.endswith('.jpg'):
self.Line_filename.setText('File Uploaded from: '+str(fileName))
img = pil.open(fileName); self.img_array = mh.stretch(np.array(img)); img.close()
self.Button_ShowImage.setEnabled(True); self.Button_Run.setEnabled(False)
# Load basic threshold params
self.T_otsu = mh.otsu(self.img_array); self.T_mean = self.img_array.mean()
else:
self.Label_status.setText('There appears to be something wrong with your input file')
self.Button_SNR.setEnabled(False); self.Check_TissueMaxima.setEnabled(False)
self.Button_DAPI.setEnabled(False);
else:
self.Label_status.setText('File not uploaded')
def chromatids_elements(TopHatedChromosome):
'''Take a High pass filtered (or top hat) image of a chromosome and label the chromatids elements
'''
threshed = TopHatedChromosome > 0
#threshed = mh.open(threshed)
labthres, _ = mh.label(threshed)
labsz = mh.labeled.labeled_size(labthres)
mh.labeled.remove_regions_where(labthres, labsz < 2, inplace=True)
threshed = labthres > 0
skel2 = mh.thin(threshed)
bp2 = branchedPoints(skel2, showSE=False) > 0
rem = np.logical_and(skel2, np.logical_not(bp2))
labskel, _ = mh.labeled.label(rem)
#print labskel.dtype
size_sk = mh.labeled.labeled_size(labskel)
#print size_sk
skelem = mh.labeled.remove_regions_where(labskel, size_sk < 4)
distances = mh.stretch(mh.distance(threshed))
surface = (distances.max() - distances)
chr_label = mh.cwatershed(surface, skelem)
#print chr_label.dtype, type(chr_label)
chr_label *= threshed
#This convertion is important !!
chr_label = chr_label.astype(np.intc)
#-------------------------------
mh.labeled.relabel(chr_label, inplace=True)
labsize2 = mh.labeled.labeled_size(chr_label)
cleaned = mh.labeled.remove_regions_where(chr_label, labsize2 < 8)
mh.labeled.relabel(cleaned, inplace=True)
return cleaned
def test_overlay():
im = mh.demos.load('luispedro', as_grey=1)
im = mh.stretch(im)
assert np.all(mh.overlay(im).max(2) == im)
edges = mh.sobel(im)
im3 = mh.overlay(im, green=edges)
assert np.all(im3[:,:,0] == im)
assert np.all(im3[:,:,2] == im)
assert np.all(im3[:,:,1] >= im )
def segment(fname):
dna = mh.imread(fname)
dna = dna[:,:,0]
sigma = 12.
dnaf = mh.gaussian_filter(dna, sigma)
T_mean = dnaf.mean()
bin_image = dnaf > T_mean
labeled, nr_objects = mh.label(bin_image)
maxima = mh.regmax(mh.stretch(dnaf))
maxima = mh.dilate(maxima, np.ones((5,5)))
maxima,_ = mh.label(maxima)
dist = mh.distance(bin_image)
dist = 255 - mh.stretch(dist)
watershed = mh.cwatershed(dist, maxima)
watershed *= bin_image
return watershed
def test_overlay():
im = mh.demos.load('luispedro', as_grey=1)
im = mh.stretch(im)
assert np.all(mh.overlay(im).max(2) == im)
edges = mh.sobel(im)
im3 = mh.overlay(im, green=edges)
assert np.all(im3[:, :, 0] == im)
assert np.all(im3[:, :, 2] == im)
assert np.all(im3[:, :, 1] >= im)
def hypersegmented_features(histone, dna, rois):
"""\
features,labels = hypersegmented_features(histone, dna, rois)
features = hypersegmented_features(histone, dna, None)
Computes hyper-segmented features on (histone/dna). If ``rois`` is not
``None``, returns the label for each region (fraction of NETs).
Parameters
----------
histone : ndarray
dna : ndarray
rois : ndarray or None
Returns
-------
features : ndarray
2D array features for each region
labels : ndarray (only if ``rois is not None``)
for the corresponding region, returns the fraction of NETs
"""
if rois is not None:
interest = rois > 0
regions, n_regions = _segment(histone)
histone = mh.stretch(histone)
dna = mh.stretch(dna)
features = []
labels = []
for ni in range(n_regions):
region = regions == (ni + 1)
if region.sum() < 16:
continue
features.append(_region_features_for(histone, dna, region))
if rois is not None:
fraction = interest[region].mean()
labels.append(fraction)
if rois is not None:
return np.array(features), np.array(labels)
return np.array(features)
def hypersegmented_features(histone, dna, rois):
'''\
features,labels = hypersegmented_features(histone, dna, rois)
features = hypersegmented_features(histone, dna, None)
Computes hyper-segmented features on (histone/dna). If ``rois`` is not
``None``, returns the label for each region (fraction of NETs).
Parameters
----------
histone : ndarray
dna : ndarray
rois : ndarray or None
Returns
-------
features : ndarray
2D array features for each region
labels : ndarray (only if ``rois is not None``)
for the corresponding region, returns the fraction of NETs
'''
if rois is not None:
interest = (rois > 0)
regions, n_regions = _segment(histone)
histone = mh.stretch(histone)
dna = mh.stretch(dna)
features = []
labels = []
for ni in range(n_regions):
region = (regions == (ni + 1))
if region.sum() < 16:
continue
features.append(_region_features_for(histone, dna, region))
if rois is not None:
fraction = interest[region].mean()
labels.append(fraction)
if rois is not None:
return np.array(features), np.array(labels)
return np.array(features)
def watershed_single_channel(np_array, threshold=10, blur_factor=9, max_bb_size=13000, min_bb_size=1000, footprint=10):
# INPUT:
# np_array: numpy image of single channel
# threshold: minimum value to be analysed
# blur_factor: gaussian blur. some blur is good. too much is bad. too little is bad too.
# max_bb_size: set maximum size for a bounding box.
# min_bb_size: set minimum size for a bounding box.
# footprint: box of size footprint x footprint used to find regions of maximum intensity.
# OUTPUT:
# return: array of bounding boxes for the channel image given
# NOTE: input values should be tuned. this alg is useless if input arguments are not optimized... right now its by hand. otherwise just leave them
# nuclear is a blured version of origional image
nuclear = mh.gaussian_filter(np_array, blur_factor)
# calculate a minimum threshold using otsu method
otsu_thresh = threshold_otsu(nuclear)
# determine minimum threshold from otsu and input argument. if little/no signal in image; otsu value can be way too low
set_thresh = None
if threshold>otsu_thresh:
set_thresh = threshold
else:
set_thresh = otsu_thresh
# set values lower than set_thresh to zero
index_otsu = nuclear < set_thresh
nuclear[index_otsu] = 0
# determine areas of maximum intensity and the distance between them
thresh = (nuclear > nuclear.mean())
dist = mh.stretch(mh.distance(thresh))
Bc = np.ones((footprint, footprint))
# the code the generate region_props from watersheding alg.
maxima = mh.morph.regmax(dist, Bc=Bc)
spots, n_spots = mh.label(maxima, Bc=Bc)
sizes = mh.labeled.labeled_size(spots)
too_big = np.where(sizes > max_bb_size)
spots = mh.labeled.remove_regions(spots, too_big)
spots = mh.labeled.remove_bordering(spots)
spots, n_left = mh.labeled.relabel(spots)
surface = (dist.max() - dist)
areas = mh.cwatershed(surface, spots)
areas *= thresh
# get list of region properties from watershed. allot of information in region_props. allot of which is inaccurate. NEVER TRUST REGIONPROPS!
region_props=regionprops(areas,intensity_image=nuclear)
# generate array of bounding boxes from measured region properties. call bbs_from_rprops()
watershed_bb_array = bbs_from_rprops(region_props, max_bb_size, min_bb_size)
return(watershed_bb_array)
def nuclei_regions(comp_map):
"""
NUCLEI_REGIONS: extract "support regions" for nuclei. This function
expects as input a "tissue components map" (as returned, for example,
by segm.tissue_components) where values of 1 indicate pixels having
a color corresponding to nuclei.
It returns a set of compact support regions corresponding to the
nuclei.
:param comp_map: numpy.ndarray
A mask identifying different tissue components, as obtained
by classification in RGB space. The value 0
See segm.tissue.tissue_components()
:return:
"""
# Deprecated:...
# img_hem, _ = rgb2he(img0, normalize=True)
# img_hem = denoise_tv_bregman(img_hem, HE_OPTS['bregm'])
# Get a mask of nuclei regions by unsupervised clustering:
# Vector Quantization: background, mid-intensity Hem and high intensity Hem
# -train the quantizer for 3 levels
# vq = KMeans(n_clusters=3)
# vq.fit(img_hem.reshape((-1,1)))
# -the level of interest is the brightest:
# k = np.argsort(vq.cluster_centers_.squeeze())[2]
# mask_hem = (vq.labels_ == k).reshape(img_hem.shape)
# ...end deprecated
# Final mask:
mask = (comp_map == 1) # use the components classified by color
# mask = morph.closing(mask, selem=HE_OPTS['strel1'])
# mask = morph.opening(mask, selem=HE_OPTS['strel1'])
# morph.remove_small_objects(mask, in_place=True)
# mask = (mask > 0)
mask = mahotas.close_holes(mask)
morph.remove_small_objects(mask, in_place=True)
dst = mahotas.stretch(mahotas.distance(mask))
Bc=np.ones((9,9))
lmax = mahotas.regmax(dst, Bc=Bc)
spots, _ = mahotas.label(lmax, Bc=Bc)
regions = mahotas.cwatershed(lmax.max() - lmax, spots) * mask
return regions
# end NUCLEI_REGIONS
def get_salt_pepper_image(self):
im = self._gray_image
salt = np.random.random(im.shape) > .975
pepper = np.random.random(im.shape) > .975
# salt is 170 & pepper is 30
# Some playing around showed that setting these to more extreme values looks
# very artificial. These look nicer
im = np.maximum(salt * 170, mh.stretch(im))
im = np.minimum(pepper * 30 + im * (~pepper), im)
return im
def _segment(cell):
# takes a numpy array of a microscopy
# segments it based on filtering the image then applying a distance transform and
# a watershed method to get the proper segmentation
import mahotas as mh
filt_cell = mh.gaussian_filter(cell, 2)
T = mh.thresholding.otsu((np.rint(filt_cell).astype('uint8')))
dist = mh.stretch(mh.distance(filt_cell > T))
Bc = np.ones((3, 3))
rmax = mh.regmin((dist))
rmax = np.invert(rmax)
labels, num_cells = mh.label(rmax, Bc)
surface = (dist.max() - dist)
areas = mh.cwatershed(dist, labels)
areas *= T
return areas
def _segment(cell):
# takes a numpy array of a microscopy
# segments it based on filtering the image then applying a distance transform and
# a watershed method to get the proper segmentation
import mahotas as mh
filt_cell = mh.gaussian_filter(cell, 2)
T = mh.thresholding.otsu((np.rint(filt_cell).astype('uint8')))
dist = mh.stretch(mh.distance(filt_cell > T))
Bc = np.ones((3,3))
rmax = mh.regmin((dist))
rmax = np.invert(rmax)
labels, num_cells = mh.label(rmax, Bc)
surface = (dist.max() - dist)
areas = mh.cwatershed(dist, labels)
areas *= T
return areas
def center_focus(im,f): r, g, b = im.transpose(2,0,1) r12 = mh.gaussian_filter(r, 12.0) g12 = mh.gaussian_filter(g, 12.0) b12 = mh.gaussian_filter(b, 12.0) im12 = mh.as_rgb(r12,g12,b12) h, w = r.shape Y, X = np.mgrid[:h,:w] Y = Y-h/2.0 Y = Y/Y.max() X = X-w/2.0 X = X/X.max() W = np.exp(-2.0*(X**2 + Y**2)/0.5) W = W - W.min() W = W/W.ptp() W = W[:, :, None] ringed = mh.stretch(im*W + (1.0-W)*im12) plt.imshow(ringed) plt.savefig(f) plt.show ()
def watershed_from_image(image):
bin_image = threshold_niblack(image)
bin_image = image > bin_image / 0.8
bin_image = mh.label(bin_image)[0]
sizes = mh.labeled.labeled_size(bin_image)
bin_image = mh.labeled.remove_regions_where(bin_image, sizes < 50)
bin_image = (bin_image > 0) * 1
distance = ndi.distance_transform_edt(bin_image)
imagef = mh.gaussian_filter(image.astype(float), 2)
maxima = mh.regmax(mh.stretch(imagef))
maxima, _ = mh.label(maxima)
markers = erode(bin_image, 1)
markers = mh.label(markers + 1)[0]
labels = sk_watershed(-distance, maxima, watershed_line=1)
watershed = labels * bin_image
return watershed
def get_Haralick(im_arr, dist=1, grid_size=1):
"""
Haralick para una imagen.
:param im_arr:
:param dist:
:param grid_size:
:return: 13*4*grid_size^2 array length
"""
img = np.asarray(im_arr).astype(int)
img = mahotas.stretch(img, 31)
window_size = (np.asarray([img.shape]) / grid_size).astype(int)[0]
im_grid = np.asarray(skimage.util.view_as_blocks(img, tuple(window_size)))
windows = []
for i in range(grid_size):
for j in range(grid_size):
windows.append(im_grid[i, j])
haralick_features = []
for i in range(len(windows)):
h = haralick(windows[i], distance=dist)
h = np.ravel(np.asarray(h))
haralick_features.append(h)
out = np.ravel(np.asarray(haralick_features))
return out
def process_image(im, d, test=False, remove_bordering=False):
plt.figure(1, frameon=False)
sigma = 75
blurred = mh.gaussian_filter(im.astype(float), sigma)
T_mean = blurred.mean()
bin_image = im > T_mean
maxima = mh.regmax(mh.stretch(blurred))
maxima, _ = mh.label(maxima)
dist = mh.distance(bin_image)
dist = 255 - mh.stretch(dist)
watershed = mh.cwatershed(dist, maxima)
_, old_nr_objects = mh.labeled.relabel(watershed)
sizes = mh.labeled.labeled_size(watershed)
min_size = 100000
filtered = mh.labeled.remove_regions_where(
watershed * bin_image, sizes < min_size)
_, nr_objects = mh.labeled.relabel(filtered)
print('Removed', old_nr_objects - nr_objects, 'small regions')
old_nr_objects = nr_objects
if (remove_bordering):
filtered = mh.labeled.remove_bordering(filtered)
labeled, nr_objects = mh.labeled.relabel(filtered)
print('Removed', old_nr_objects - nr_objects, 'bordering cells')
print("Number of cells: {}".format(nr_objects))
fin_weights = mh.labeled_sum(im.astype(np.uint32), labeled)
fin_sizes = mh.labeled.labeled_size(labeled)
fin_result = fin_weights / fin_sizes
if (test):
f, axarr = plt.subplots(2, 2)
for i in range(2):
for j in range(2):
axarr[i][j].axis('off')
axarr[0, 0].imshow(im)
axarr[0, 0].set_title('Source')
axarr[0, 1].imshow(labeled)
axarr[0, 1].set_title('Labeled')
axarr[1, 0].imshow(watershed)
axarr[1, 0].set_title('Watershed')
axarr[1, 1].imshow(blurred)
axarr[1, 1].set_title('Blurred')
for i in range(1, nr_objects + 1):
print("Cell {} average luminescence is {}".format(
i, fin_result[i]))
bbox = mh.bbox((labeled == i))
plt.text((bbox[2] + bbox[3]) / 2, (bbox[0] + bbox[1]
) / 2, str(i), fontsize=20, color='black')
# plt.show()
plt.savefig("test" + str(nr_objects) + ".svg",
format='svg', bbox_inches='tight', dpi=1200)
else:
for i in range(1, nr_objects + 1):
bbox = mh.bbox((labeled == i))
cell = (im * (labeled == i))[bbox[0]:bbox[1], bbox[2]:bbox[3]]
hashed = hashlib.sha1(im).hexdigest()
imsave(d + data_dir + hashed + '-' + str(i) +
'.png', imresize(cell, (img_rows, img_cols)))
import mahotas as mh
from jugfile import method1
from matplotlib import cm
import numpy as np
im = mh.imread('images/dna-21.jpg')
mh.imsave('image_stretched.jpeg', mh.stretch(im.astype(float)**.01))
m1 = method1.f('images/dna-21.jpg', 2)
m1 = m1.astype(np.uint8)
color = ((cm.rainbow(m1.astype(float)/m1.max())[:,:,:3]).reshape(m1.shape+(3,)))
color[m1 == 0] = (0,0,0)
mh.imsave('image_method1.jpeg', mh.stretch(color))
ref = mh.imread('references/dna-21.png')
color = ((cm.rainbow(ref.astype(float)/ref.max())[:,:,:3]).reshape(m1.shape+(3,)))
color[ref == 0] = (0,0,0)
mh.imsave('image_reference.jpeg', mh.stretch(color))
from imread import imread
import numpy as np
import mahotas as mh
lenna = imread('../DATA/Lenna.png', as_grey=True)
views = []
view_titles = []
views.append(lenna)
view_titles.append('Lenna, grayscale')
salt = np.random.random(lenna.shape) > .975
pepper = np.random.random(lenna.shape) > .975
lenna = mh.stretch(lenna)
lenna = np.maximum(salt*170, lenna)
lenna = np.minimum(pepper*30 + lenna*(~pepper), lenna)
views.append(lenna)
view_titles.append('Lenna, salt & pepper')
# Initial view
index = 0
f = Figure()
a = f.add_subplot(111)
a.imshow(views[index], cmap = cm.Greys_r)
a.set_title(view_titles[index])
#/\#/\#
import mahotas
import numpy as np
from matplotlib import pyplot as plt
import random
from matplotlib import colors as c
nuclear = mahotas.imread('data/flower.png')
nuclear = nuclear[:, :, 0]
nuclear = mahotas.gaussian_filter(nuclear, 1.)
threshed = (nuclear > nuclear.mean())
distances = mahotas.stretch(mahotas.distance(threshed))
Bc = np.ones((9, 9))
maxima = mahotas.morph.regmax(distances, Bc=Bc)
spots, n_spots = mahotas.label(maxima, Bc=Bc)
surface = (distances.max() - distances)
areas = mahotas.cwatershed(surface, spots)
areas *= threshed
plt.jet()
rmap = c.ListedColormap(np.random.rand(256, 3))
plt.imshow(areas, cmap=rmap)
plt.show()
def main(image_1, image_2, weight_1, weight_2, plot=False):
'''Combines `image_1` with `image_2`.
Parameters
----------
input_mask_1: numpy.ndarray[numpy.uint8 or numpy.uint16]
2D unsigned integer array
input_mask_2: numpy.ndarray[numpy.uint8 or numpy.uint16]
2D unsigned integer array
weight_1: int
weight for `image_1`
weight_2: int
weight for `image_2`
Returns
-------
jtmodules.combine_channels.Output
Raises
------
ValueError
when `weight_1` or `weight_2` are not positive integers
ValueError
when `image_1` and `image_2` don't have the same dimensions
and data type and if they don't have unsigned integer type
'''
if not isinstance(weight_1, int):
raise TypeError('Weight #1 must have integer type.')
if not isinstance(weight_2, int):
raise TypeError('Weight #2 must have integer type.')
if weight_1 < 1:
raise ValueError('Weight #1 must be a positive integer.')
if weight_2 < 1:
raise ValueError('Weight #2 must be a positive integer.')
logger.info('weight for first image: %d', weight_1)
logger.info('weight for second image: %d', weight_2)
if image_1.shape != image_2.shape:
raise ValueError('The two images must have identical dimensions.')
if image_1.dtype != image_2.dtype:
raise ValueError('The two images must have identical data type.')
if image_1.dtype == np.uint8:
max_val = 2**8 - 1
elif image_1.dtype == np.uint16:
max_val = 2**16 - 1
else:
raise ValueError('The two images must have unsigned integer type.')
logger.info('cast images to type float for arythmetics')
img_1 = mh.stretch(image_1, 0, 1, float)
img_2 = mh.stretch(image_2, 0, 1, float)
logger.info('combine images using the provided weights')
combined_image = img_1 * weight_1 + img_2 * weight_2
logger.info('cast combined image back to correct data type')
combined_image = mh.stretch(combined_image, 0, max_val, image_1.dtype)
if plot:
from jtlib import plotting
plots = [
plotting.create_intensity_image_plot(image_1, 'ul'),
plotting.create_intensity_image_plot(image_2, 'ur'),
plotting.create_intensity_image_plot(combined_image, 'll')
]
figure = plotting.create_figure(plots, title='combined image')
else:
figure = str()
return Output(combined_image, figure)
def region_growing(labelImg):
distances = mahotas.stretch(mahotas.distance(labelImg > 0))
surface = numpy.int32(distances.max() - distances)
areas = mahotas.cwatershed(surface, labelImg)
return areas
def check_sigma(sigma):
dnaf = mh.gaussian_filter(dna.astype(float), sigma)
maxima = mh.regmax(mh.stretch(dnaf))
maxima = mh.dilate(maxima, np.ones((5,5)))
plt.imshow(mh.as_rgb(np.maximum(255*maxima, dnaf), dnaf, dna > T_mean))
import mahotas as mh
import numpy as np
im = mh.imread('lenna.jpg')
r,g,b = im.transpose(2,0,1)
h,w = r.shape
r12 = mh.gaussian_filter(r, 12.)
g12 = mh.gaussian_filter(g, 12.)
b12 = mh.gaussian_filter(b, 12.)
im12 = mh.as_rgb(r12,g12,b12)
X,Y = np.mgrid[:h,:w]
X = X-h/2.
Y = Y-w/2.
X /= X.max()
Y /= Y.max()
C = np.exp(-2.*(X**2+ Y**2))
C -= C.min()
C /= C.ptp()
C = C[:,:,None]
ring = mh.stretch(im*C + (1-C)*im12)
mh.imsave('lenna-ring.jpg', ring)
def extract(self):
'''Extracts Gabor texture features by filtering the intensity image with
Gabor kernels for a defined range of `frequency` and `theta` values and
then calculating a score for each object.
Returns
-------
pandas.DataFrame
extracted feature values for each object in `label_image`
'''
# Create an empty dataset in case no objects were detected
logger.info('extract texture features')
features = list()
for obj in self.object_ids:
mask = self.get_object_mask_image(obj)
label = mask.astype(np.int32)
img = self.get_object_intensity_image(obj)
img[~mask] = 0
values = list()
# Gabor
logger.debug('extract Gabor features for object #%d', obj)
for freq in self.frequencies:
best_score = 0
for angle in range(self.theta_range):
theta = np.pi * angle / self.theta_range
g = gabor(img, label, freq, theta)
score_r = ndi.measurements.sum(
g.real, label, np.arange(1, dtype=np.int32) + 1
)
score_i = ndi.measurements.sum(
g.imag, label, np.arange(1, dtype=np.int32) + 1
)
score = np.sqrt(score_r**2 + score_i**2)
best_score = np.max([best_score, score])
values.append(best_score)
# Threshold Adjacency Statistics
logger.debug('extract TAS features for object #%d', obj)
tas_values = mh.features.pftas(img, T=self._threshold)
values.extend(tas_values)
# Hu
logger.debug('extract Hu moments for object #%d', obj)
region = self.object_properties[obj]
hu_values = region.weighted_moments_hu
values.extend(hu_values)
# Local Binary Pattern
logger.debug('extract Local Binary Patterns for object #%d', obj)
for r in self.radius:
# We may want to use more points, but the number of features
# increases exponentially with the number of neighbourhood
# points.
vals = mh.features.lbp(img, radius=r, points=8)
values.extend(vals)
if self.compute_haralick:
# Haralick
logger.debug('extract Haralick features for object #%d', obj)
# NOTE: Haralick features are computed on 8-bit images.
clipped_img = np.clip(img, 0, self._clip_value)
rescaled_img = mh.stretch(clipped_img)
haralick_values = mh.features.haralick(
img, ignore_zeros=False, return_mean=True
)
if not isinstance(haralick_values, np.ndarray):
# NOTE: setting `ignore_zeros` to True creates problems for some
# objects, when all values of the adjacency matrices are zeros
haralick_values = np.empty((len(self.names), ), dtype=float)
haralick_values[:] = np.NAN
values.extend(haralick_values)
features.append(values)
return pd.DataFrame(features, columns=self.names, index=self.object_ids)
bin_image = dnaf > T_mean
plt.imshow(bin_image)
labeled, nr_objects = mh.label(bin_image)
print(nr_objects)
plt.imshow(labeled)
plt.jet()
@interact(sigma=(1.,16.))
def check_sigma(sigma):
dnaf = mh.gaussian_filter(dna.astype(float), sigma)
maxima = mh.regmax(mh.stretch(dnaf))
maxima = mh.dilate(maxima, np.ones((5,5)))
plt.imshow(mh.as_rgb(np.maximum(255*maxima, dnaf), dnaf, dna > T_mean))
sigma = 12.0
dnaf = mh.gaussian_filter(dna.astype(float),sigma)
maxima = mh.regmax(mh.stretch(dnaf))
maxima,_= mh.label(maxima)
plt.imshow(maxima)
dist = mh.distance(bin_image)
plt.imshow(dist)
dist = 255 - mh.stretch(dist)
watershed = mh.cwatershed(dist,maxima)
plt.imshow(watershed)
watershed *= bin_image
plt.imshow(watershed)
watershed = mh.labeled.remove_bordering(watershed)
plt.imshow(watershed)
sizes = mh.labeled.labeled_size(watershed)
# The conversion below is not necessary in newer versions of mahotas: watershed = watershed.astype(np.intc)
@interact(min_size=(100,4000,20))
def do_plot(min_size):
def run(self, workspace):
x_name = self.x_name.value
y_name = self.y_name.value
images = workspace.image_set
x = images.get_image(x_name)
dimensions = x.dimensions
x_data = x.pixel_data
if self.operation.value == "Distance":
original_shape = x_data.shape
factor = self.downsample.value
if factor > 1:
if x.volumetric:
factors = (1, factor, factor)
else:
factors = (factor, factor)
x_data = skimage.transform.downscale_local_mean(
x_data, factors)
threshold = skimage.filters.threshold_otsu(x_data)
x_data = x_data > threshold
distance = scipy.ndimage.distance_transform_edt(x_data)
distance = mahotas.stretch(distance)
surface = distance.max() - distance
if x.volumetric:
footprint = numpy.ones(
(self.connectivity.value, self.connectivity.value,
self.connectivity.value))
else:
footprint = numpy.ones(
(self.connectivity.value, self.connectivity.value))
peaks = mahotas.regmax(distance, footprint)
if x.volumetric:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16, 16)))
else:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16)))
y_data = mahotas.cwatershed(surface, markers)
y_data = y_data * x_data
if factor > 1:
y_data = skimage.transform.resize(y_data,
original_shape,
mode="edge",
order=0,
preserve_range=True)
y_data = numpy.rint(y_data).astype(numpy.uint16)
else:
markers_name = self.markers_name.value
markers = images.get_image(markers_name)
markers_data = markers.pixel_data
if x.multichannel:
x_data = skimage.color.rgb2gray(x_data)
if markers.multichannel:
markers_data = skimage.color.rgb2gray(markers_data)
mask_data = None
if not self.mask_name.is_blank:
mask_name = self.mask_name.value
mask = images.get_image(mask_name)
mask_data = mask.pixel_data
y_data = skimage.morphology.watershed(image=x_data,
markers=markers_data,
mask=mask_data)
y_data = skimage.measure.label(y_data)
objects = cellprofiler.object.Objects()
objects.segmented = y_data
objects.parent_image = x
workspace.object_set.add_objects(objects, y_name)
self.add_measurements(workspace)
if self.show_window:
workspace.display_data.x_data = x.pixel_data
workspace.display_data.y_data = y_data
workspace.display_data.dimensions = dimensions
def test_run_distance_declump_intensity(
image, module, image_set, workspace, connectivity, compactness, watershed_line
):
module.use_advanced.value = True
module.operation.value = "Distance"
module.x_name.value = "binary"
module.y_name.value = "watershed"
module.connectivity.value = connectivity
module.footprint.value = 3
data = image.pixel_data
if image.multichannel:
data = skimage.color.rgb2gray(data)
threshold = skimage.filters.threshold_otsu(data)
binary = data > threshold
image_set.add(
"binary",
cellprofiler_core.image.Image(
image=binary, convert=False, dimensions=image.dimensions
),
)
module.declump_method.value = "Intensity"
module.reference_name.value = "gradient"
module.gaussian_sigma.value = 1
# must pass pixel data into image set for intensity declumping
gradient = image.pixel_data
image_set.add(
"gradient",
cellprofiler_core.image.Image(
image=gradient, convert=False, dimensions=image.dimensions
),
)
# set the structuring element, used for declumping
if image.dimensions == 3:
module.structuring_element.value = "Ball,1"
selem = skimage.morphology.ball(1)
else:
module.structuring_element.value = "Disk,1"
selem = skimage.morphology.disk(1)
# run the module
module.run(workspace)
# distance-based watershed
distance = scipy.ndimage.distance_transform_edt(binary)
distance = mahotas.stretch(distance)
surface = distance.max() - distance
if image.volumetric:
footprint = numpy.ones((3, 3, 3))
else:
footprint = numpy.ones((3, 3))
peaks = mahotas.regmax(distance, footprint)
if image.volumetric:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16, 16)))
else:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16)))
watershed_distance = mahotas.cwatershed(surface, markers)
watershed_distance = watershed_distance * binary
# intensity-based declumping
peak_image = scipy.ndimage.distance_transform_edt(watershed_distance > 0)
# Set the image as a float and rescale to full bit depth
watershed_image = skimage.img_as_float(gradient, force_copy=True)
watershed_image -= watershed_image.min()
watershed_image = 1 - watershed_image
if image.multichannel:
watershed_image = skimage.color.rgb2gray(watershed_image)
watershed_image = skimage.filters.gaussian(watershed_image, sigma=module.gaussian_sigma.value)
seed_coords = skimage.feature.peak_local_max(peak_image,
min_distance=module.min_dist.value,
threshold_rel=module.min_intensity.value,
exclude_border=module.exclude_border.value,
num_peaks=module.max_seeds.value if module.max_seeds.value != -1
else numpy.inf)
seeds = numpy.zeros_like(peak_image, dtype=bool)
seeds[tuple(seed_coords.T)] = True
seeds = skimage.morphology.binary_dilation(seeds, selem)
number_objects = skimage.measure.label(watershed_distance, return_num=True)[1]
seeds_dtype = (numpy.uint16 if number_objects < numpy.iinfo(numpy.uint16).max else numpy.uint32)
seeds = scipy.ndimage.label(seeds)[0]
markers = numpy.zeros_like(seeds, dtype=seeds_dtype)
markers[seeds > 0] = -seeds[seeds > 0]
expected = skimage.segmentation.watershed(
connectivity=connectivity,
image=watershed_image,
markers=markers,
mask=binary !=0
)
zeros = numpy.where(expected==0)
expected += numpy.abs(numpy.min(expected)) + 1
expected[zeros] = 0
expected = skimage.measure.label(expected)
actual = workspace.get_objects("watershed")
numpy.testing.assert_array_equal(expected, actual.segmented)
def separate_clumped_objects(clumps_image, min_cut_area, min_area, max_area,
max_circularity, max_convexity):
'''Separates objects in `clumps_image` based on morphological criteria.
Parameters
----------
clumps_image: numpy.ndarray[Union[numpy.int32, numpy.bool]]
objects that should be separated
min_cut_area: int
minimal area an object must have (prevents cuts that would result
in too small objects)
min_area: int
minimal area an object must have to be considered a clump
max_area: int
maximal area an object can have to be considered a clump
max_circularity: float
maximal circularity an object must have to be considerd a clump
max_convexity: float
maximal convexity an object must have to be considerd a clump
Returns
-------
numpy.ndarray[numpy.uint32]
separated objects
See also
--------
:class:`jtlib.features.Morphology`
'''
logger.info('separate clumped objects')
label_image, n_objects = mh.label(clumps_image)
if n_objects == 0:
logger.debug('no objects')
return label_image
pad = 1
cutting_pass = 1
separated_image = label_image.copy()
while True:
logger.info('cutting pass #%d', cutting_pass)
cutting_pass += 1
label_image = mh.label(label_image > 0)[0]
f = Morphology(label_image)
values = f.extract()
index = (
(min_area < values['Morphology_Area']) &
(values['Morphology_Area'] <= max_area) &
(values['Morphology_Convexity'] <= max_convexity) &
(values['Morphology_Circularity'] <= max_circularity)
)
clumped_ids = values[index].index.values
not_clumped_ids = values[~index].index.values
if len(clumped_ids) == 0:
logger.debug('no more clumped objects')
break
mh.labeled.remove_regions(label_image, not_clumped_ids, inplace=True)
mh.labeled.relabel(label_image, inplace=True)
bboxes = mh.labeled.bbox(label_image)
for oid in np.unique(label_image[label_image > 0]):
bbox = bboxes[oid]
logger.debug('process clumped object #%d', oid)
obj_image = extract_bbox(label_image, bboxes[oid], pad=pad)
obj_image = obj_image == oid
# Rescale distance intensities to make them independent of clump size
dist = mh.stretch(mh.distance(obj_image))
# Find peaks that can be used as seeds for the watershed transform
thresh = mh.otsu(dist)
peaks = dist > thresh
n = mh.label(peaks)[1]
if n == 1:
logger.debug(
'only one peak detected - perform iterative erosion'
)
# Iteratively shrink the peaks until we have two peaks that we
# can use to separate the clump.
while True:
tmp = mh.morph.open(mh.morph.erode(peaks))
n = mh.label(tmp)[1]
if n == 2 or n == 0:
if n == 2:
peaks = tmp
break
peaks = tmp
# Select the two biggest peaks, since we want only two objects.
peaks = mh.label(peaks)[0]
sizes = mh.labeled.labeled_size(peaks)
index = np.argsort(sizes)[::-1][1:3]
for label in np.unique(peaks):
if label not in index:
peaks[peaks == label] = 0
peaks = mh.labeled.relabel(peaks)[0]
regions = mh.cwatershed(np.invert(dist), peaks)
# Use the line separating watershed regions to make the cut
se = np.ones((3,3), np.bool)
line = mh.labeled.borders(regions, Bc=se)
line[~obj_image] = 0
line = mh.morph.dilate(line)
# Ensure that cut is reasonable given user-defined criteria
test_cut_image = obj_image.copy()
test_cut_image[line] = False
subobjects, n_subobjects = mh.label(test_cut_image)
sizes = mh.labeled.labeled_size(subobjects)
smaller_object_area = np.min(sizes)
smaller_id = np.where(sizes == smaller_object_area)[0][0]
smaller_object = subobjects == smaller_id
do_cut = (
(smaller_object_area > min_cut_area) &
(np.sum(line) > 0)
)
if do_cut:
logger.debug('cut object #%d', oid)
y, x = np.where(line)
y_offset, x_offset = bboxes[oid][[0, 2]] - pad - 1
y += y_offset
x += x_offset
label_image[y, x] = 0
separated_image[y, x] = 0
else:
logger.debug('don\'t cut object #%d', oid)
mh.labeled.remove_regions(label_image, oid, inplace=True)
return mh.label(separated_image)[0]
# This code is supporting material for the book
# Building Machine Learning Systems with Python
# by Willi Richert and Luis Pedro Coelho
# published by PACKT Publishing
#
# It is made available under the MIT License
from matplotlib import pyplot as plt
import numpy as np
import mahotas as mh
image = mh.imread('../1400OS_10_01.jpeg')
image = mh.colors.rgb2gray(image, dtype=np.uint8)
image = image[::4, ::4]
thresh = mh.sobel(image)
filtered = mh.sobel(image, just_filter=True)
thresh = mh.dilate(thresh, np.ones((7, 7)))
filtered = mh.dilate(mh.stretch(filtered), np.ones((7, 7)))
h, w = thresh.shape
canvas = 255 * np.ones((h, w * 2 + 64), np.uint8)
canvas[:, :w] = thresh * 255
canvas[:, -w:] = filtered
mh.imsave('../1400OS_10_09+.jpg', canvas)
def features_for(imfile):
return mh.features.haralick(mh.stretch(mh.imread(imfile))//4).mean(0)[[1,12]]
# This code is supporting material for the book
# Building Machine Learning Systems with Python
# by Willi Richert and Luis Pedro Coelho
# published by PACKT Publishing
#
# It is made available under the MIT License
from matplotlib import pyplot as plt
import numpy as np
import mahotas as mh
image = mh.imread('../1400OS_10_01.jpeg')
image = mh.colors.rgb2gray(image, dtype=np.uint8)
image = image[::4, ::4]
thresh = mh.sobel(image)
filtered = mh.sobel(image, just_filter=True)
thresh = mh.dilate(thresh, np.ones((7, 7)))
filtered = mh.dilate(mh.stretch(filtered), np.ones((7, 7)))
h, w = thresh.shape
canvas = 255 * np.ones((h, w * 2 + 64), np.uint8)
canvas[:, :w] = thresh * 255
canvas[:, -w:] = filtered
mh.imsave('../1400OS_10_09+.jpg', canvas)
sizes = mahotas.labeled.labeled_size(colonies)
# print(sizes)
too_small = np.where(sizes < 100)
colonies = mahotas.labeled.remove_regions(colonies, too_small)
#colonies = mahotas.labeled.remove_bordering(colonies)
colonies, n_colonies = mahotas.labeled.relabel(colonies)
print('Found {} colonies.'.format(n_colonies))
# plt.imshow(colonies)
# print(colonies)
# Investigate nuclei within cell clusters
# Now, we compute the distance transform:
distances = mahotas.stretch(mahotas.distance(local_threshed))
# We find and label the regional maxima:
Bc = np.ones((9,9))
maxima = mahotas.morph.regmax(distances, Bc=Bc)
spots,n_spots = mahotas.label(maxima, Bc=Bc)
print('Found {} maxima.'.format(n_spots))
# plt.imshow(spots)
# Finally, to obtain the image above, we invert the distance transform
# (because of the way that cwatershed is defined) and compute the watershed:
surface = (distances.max() - distances)
areas = mahotas.cwatershed(surface, spots)
areas *= local_threshed
import numpy as np
from matplotlib import pyplot as plt
try:
nuclear_path = path.join(
path.dirname(path.abspath(__file__)),
'data',
'nuclear.png')
except NameError:
nuclear_path = path.join('data', 'nuclear.png')
nuclear = mahotas.imread(nuclear_path)
nuclear = nuclear[:,:,0]
nuclear = mahotas.gaussian_filter(nuclear, 1.)
threshed = (nuclear > nuclear.mean())
distances = mahotas.stretch(mahotas.distance(threshed))
Bc = np.ones((9,9))
maxima = mahotas.morph.regmax(distances, Bc=Bc)
spots,n_spots = mahotas.label(maxima, Bc=Bc)
surface = (distances.max() - distances)
areas = mahotas.cwatershed(surface, spots)
areas *= threshed
import random
from matplotlib import colors as c
colors = map(plt.cm.jet,range(0, 256, 4))
random.shuffle(colors)
colors[0] = (0.,0.,0.,1.)
def region_growing(labelImg):
distances = mahotas.stretch(mahotas.distance(labelImg>0))
surface = numpy.int32(distances.max() - distances)
areas = mahotas.cwatershed(surface, labelImg)
return areas
im = mh.imread('lenna.jpg')
# This breaks up the image into RGB channels
r, g, b = im.transpose(2, 0, 1)
h, w = r.shape
# smooth the image per channel:
r12 = mh.gaussian_filter(r, 12.)
g12 = mh.gaussian_filter(g, 12.)
b12 = mh.gaussian_filter(b, 12.)
# build back the RGB image
im12 = mh.as_rgb(r12, g12, b12)
X, Y = np.mgrid[:h, :w]
X = X - h / 2.
Y = Y - w / 2.
X /= X.max()
Y /= Y.max()
# Array C will have the highest values in the center, fading out to the edges:
C = np.exp(-2. * (X**2 + Y**2))
C -= C.min()
C /= C.ptp()
C = C[:, :, None]
# The final result is sharp in the centre and smooths out to the borders:
ring = mh.stretch(im * C + (1 - C) * im12)
mh.imsave('lenna-ring.jpg', ring)
def separate_clumped_objects(clumps_image, min_cut_area, min_area, max_area,
max_circularity, max_convexity):
'''Separates objects in `clumps_image` based on morphological criteria.
Parameters
----------
clumps_image: numpy.ndarray[Union[numpy.int32, numpy.bool]]
objects that should be separated
min_cut_area: int
minimal area an object must have (prevents cuts that would result
in too small objects)
min_area: int
minimal area an object must have to be considered a clump
max_area: int
maximal area an object can have to be considered a clump
max_circularity: float
maximal circularity an object must have to be considerd a clump
max_convexity: float
maximal convexity an object must have to be considerd a clump
Returns
-------
numpy.ndarray[numpy.uint32]
separated objects
See also
--------
:class:`jtlib.features.Morphology`
'''
logger.info('separate clumped objects')
label_image, n_objects = mh.label(clumps_image)
if n_objects == 0:
logger.debug('no objects')
return label_image
pad = 1
cutting_pass = 1
separated_image = label_image.copy()
while True:
logger.info('cutting pass #%d', cutting_pass)
cutting_pass += 1
label_image = mh.label(label_image > 0)[0]
f = Morphology(label_image)
values = f.extract()
index = (
(min_area < values['Morphology_Area']) &
(values['Morphology_Area'] <= max_area) &
(values['Morphology_Convexity'] <= max_convexity) &
(values['Morphology_Circularity'] <= max_circularity)
)
clumped_ids = values[index].index.values
not_clumped_ids = values[~index].index.values
if len(clumped_ids) == 0:
logger.debug('no more clumped objects')
break
mh.labeled.remove_regions(label_image, not_clumped_ids, inplace=True)
mh.labeled.relabel(label_image, inplace=True)
bboxes = mh.labeled.bbox(label_image)
for oid in np.unique(label_image[label_image > 0]):
bbox = bboxes[oid]
logger.debug('process clumped object #%d', oid)
obj_image = extract_bbox(label_image, bboxes[oid], pad=pad)
obj_image = obj_image == oid
# Rescale distance intensities to make them independent of clump size
dist = mh.stretch(mh.distance(obj_image))
# Find peaks that can be used as seeds for the watershed transform
thresh = mh.otsu(dist)
peaks = dist > thresh
n = mh.label(peaks)[1]
if n == 1:
logger.debug(
'only one peak detected - perform iterative erosion'
)
# Iteratively shrink the peaks until we have two peaks that we
# can use to separate the clump.
while True:
tmp = mh.morph.open(mh.morph.erode(peaks))
n = mh.label(tmp)[1]
if n == 2 or n == 0:
if n == 2:
peaks = tmp
break
peaks = tmp
# Select the two biggest peaks, since we want only two objects.
peaks = mh.label(peaks)[0]
sizes = mh.labeled.labeled_size(peaks)
index = np.argsort(sizes)[::-1][1:3]
for label in np.unique(peaks):
if label not in index:
peaks[peaks == label] = 0
peaks = mh.labeled.relabel(peaks)[0]
regions = mh.cwatershed(np.invert(dist), peaks)
# Use the line separating watershed regions to make the cut
se = np.ones((3,3), np.bool)
line = mh.labeled.borders(regions, Bc=se)
line[~obj_image] = 0
line = mh.morph.dilate(line)
# Ensure that cut is reasonable given user-defined criteria
test_cut_image = obj_image.copy()
test_cut_image[line] = False
subobjects, n_subobjects = mh.label(test_cut_image)
sizes = mh.labeled.labeled_size(subobjects)
smaller_object_area = np.min(sizes)
smaller_id = np.where(sizes == smaller_object_area)[0][0]
smaller_object = subobjects == smaller_id
do_cut = (
(smaller_object_area > min_cut_area) &
(np.sum(line) > 0)
)
if do_cut:
logger.debug('cut object #%d', oid)
y, x = np.where(line)
y_offset, x_offset = bboxes[oid][[0, 2]] - pad - 1
y += y_offset
x += x_offset
label_image[y, x] = 0
separated_image[y, x] = 0
else:
logger.debug('don\'t cut object #%d', oid)
mh.labeled.remove_regions(label_image, oid, inplace=True)
return mh.label(separated_image)[0]
plt.show()
# Gaussian Filter
im8 = mh.gaussian_filter(image,8)
plt.imshow(im8)
plt.savefig('Lenna_gaussian_filter_8.jpg')
plt.show()
# Salt and Pepper noise
image = mh.colors.rgb2gray(image)
plt.gray()
salt = np.random.random(image.shape) >.8
pepper = np.random.random(image.shape) >.8
image = mh.stretch(image)
image = np.maximum(salt*170, image)
image = np.minimum(pepper*30+image* (~pepper), image)
plt.imshow(image)
plt.savefig('Lenna_salt_pepper.jpg')
plt.show ()
# Putting center in focus
def center_focus(im,f):
r, g, b = im.transpose(2,0,1)
r12 = mh.gaussian_filter(r, 12.0)
g12 = mh.gaussian_filter(g, 12.0)
b12 = mh.gaussian_filter(b, 12.0)
im12 = mh.as_rgb(r12,g12,b12)
# This code is supporting material for the book
# Building Machine Learning Systems with Python
# by Willi Richert and Luis Pedro Coelho
# published by PACKT Publishing
#
# It is made available under the MIT License
import mahotas as mh
from mahotas.colors import rgb2grey
import numpy as np
im = mh.imread('lenna.jpg')
im = rgb2grey(im)
salt = np.random.random(im.shape) > .975
pepper = np.random.random(im.shape) > .975
im = np.maximum(salt * 170, mh.stretch(im))
im = np.minimum(pepper * 30 + im * (~pepper), im)
mh.imsave('../1400OS_10_13+.jpg', im.astype(np.uint8))
import numpy as np
import mahotas as mh
image = mh.imread('../SimpleImageDataset/building05.jpg')
image = mh.colors.rgb2gray(image)
# Compute Gaussian filtered versions with increasing kernel widths
im8 = mh.gaussian_filter(image, 8)
im16 = mh.gaussian_filter(image, 16)
im32 = mh.gaussian_filter(image, 32)
# We now build a composite image with three panels:
#
# [ IM8 | | IM16 | | IM32 ]
h, w = im8.shape
canvas = np.ones((h, 3 * w + 256), np.uint8)
canvas *= 255
canvas[:, :w] = im8
canvas[:, w + 128:2 * w + 128] = im16
canvas[:, -w:] = im32
mh.imsave('../1400OS_10_05+.jpg', canvas[:, ::2])
# Threshold the image
# We need to first stretch it to convert to an integer image
im32 = mh.stretch(im32)
ot32 = mh.otsu(im32)
# Convert to 255 np.uint8 to match the other images
im255 = 255 * (im32 > ot32).astype(np.uint8)
mh.imsave('../1400OS_10_06+.jpg', im255)
def run(self, workspace):
x_name = self.x_name.value
y_name = self.y_name.value
images = workspace.image_set
x = images.get_image(x_name)
dimensions = x.dimensions
x_data = x.pixel_data
if self.operation.value == "Distance":
original_shape = x_data.shape
if x.volumetric:
x_data = skimage.transform.resize(x_data, (original_shape[0], 256, 256), order=0, mode="edge")
distance = scipy.ndimage.distance_transform_edt(x_data)
distance = mahotas.stretch(distance)
surface = distance.max() - distance
if x.volumetric:
footprint = numpy.ones((self.connectivity.value, self.connectivity.value, self.connectivity.value))
else:
footprint = numpy.ones((self.connectivity.value, self.connectivity.value))
peaks = mahotas.regmax(distance, footprint)
if x.volumetric:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16, 16)))
else:
markers, _ = mahotas.label(peaks, numpy.ones((16, 16)))
y_data = mahotas.cwatershed(surface, markers)
y_data = y_data * x_data
if x.volumetric:
y_data = skimage.transform.resize(y_data, original_shape, order=0, mode="edge")
else:
markers_name = self.markers_name.value
markers = images.get_image(markers_name)
data = x_data
markers_data = markers.pixel_data
mask_data = None
if not self.mask_name.is_blank:
mask_name = self.mask_name.value
mask = images.get_image(mask_name)
mask_data = mask.pixel_data
y_data = skimage.morphology.watershed(
image=data,
markers=markers_data,
mask=mask_data
)
y_data = skimage.measure.label(y_data)
objects = cellprofiler.object.Objects()
objects.segmented = y_data
objects.parent_image = x
workspace.object_set.add_objects(objects, y_name)
self.add_measurements(workspace.measurements, y_data)
if self.show_window:
workspace.display_data.x_data = x.pixel_data
workspace.display_data.y_data = y_data
workspace.display_data.dimensions = dimensions
def getchannel(channel):
if type(self.channels['protein']) == list:
orig=self.channeldata[channel][idx,:,:]
else:
orig=self.channeldata[channel]
return mahotas.stretch(orig)
def overlay_thresh(im):
return overlay_withcolor(im, stretch(im) < double_otsu_smaller(stretch(im)))
