def placePointsInVolumen(points, shape):
volumen = numpy.zeros(shape)
points = numpy.maximum(points, numpy.array((0, 0, 0)))
points = numpy.minimum(points, numpy.array(shape) - 1)
for point in (numpy.floor(points)).astype(int):
volumen[point[0], point[1], point[2]] = 1
return volumen
def placePointsInVolumen(points, shape):
volumen = numpy.zeros(shape)
points = numpy.maximum(points, numpy.array((0, 0, 0)))
points = numpy.minimum(points, numpy.array(shape) - 1)
for point in (numpy.floor(points)).astype(int):
volumen[point[0], point[1], point[2]] = 1
return volumen
def pmapWatershed(pmap, raw, visu=True, seedThreshold=0.6):
viewGrayData = [(pmap, "pmap") ]
viewLabelData= []
print "densoise"
pmapD = denoise.tvBregman(pmap, weight=4.5, isotropic=True).astype(numpy.float32)
pmapG = vigra.filters.gaussianSmoothing(pmap, 1.0)
viewGrayData.append((pmapG, "pmapGauss"))
viewGrayData.append((pmapD, "pmapTotalVariation"))
viewGrayData.append((raw, "raw"))
#addHocViewer(viewGrayData, viewLabelData, visu=visu)
print "compute local minima "
localMin = vigra.analysis.extendedLocalMinima3D(pmapD,neighborhood=26)
localMin2 = localMin.astype(numpy.float32)
print "tweak min"
localMin2 *= pmap
whereZero = numpy.where(localMin == 0)
localMin2[whereZero] = 100.0
whereMin = numpy.where(localMin2 <= seedThreshold)
filteredLocalMin = numpy.zeros(localMin.shape, dtype=numpy.uint8)
filteredLocalMin[whereMin] = 1
viewGrayData.append([localMin,"localMin"])
viewGrayData.append([filteredLocalMin,"filteredLocalMin"])
# compute connected components
seeds = vigra.analysis.labelVolumeWithBackground(filteredLocalMin, neighborhood=26)
viewLabelData.append([seeds, "seeds"])
print "watersheds"
seg, nseg = vigra.analysis.watersheds(pmapG.astype(numpy.float32), seeds=seeds.astype(numpy.uint32))
print "nseg",nseg
viewLabelData.append([seg, "seg"])
addHocViewer(viewGrayData, viewLabelData, visu=visu)
return se
def pmapWatershed(pmap, raw, visu=True, seedThreshold=0.6):
viewGrayData = [(pmap, "pmap")]
viewLabelData = []
print "densoise"
pmapD = denoise.tvBregman(pmap, weight=4.5,
isotropic=True).astype(numpy.float32)
pmapG = vigra.filters.gaussianSmoothing(pmap, 1.0)
viewGrayData.append((pmapG, "pmapGauss"))
viewGrayData.append((pmapD, "pmapTotalVariation"))
viewGrayData.append((raw, "raw"))
#addHocViewer(viewGrayData, viewLabelData, visu=visu)
print "compute local minima "
localMin = vigra.analysis.extendedLocalMinima3D(pmapD, neighborhood=26)
localMin2 = localMin.astype(numpy.float32)
print "tweak min"
localMin2 *= pmap
whereZero = numpy.where(localMin == 0)
localMin2[whereZero] = 100.0
whereMin = numpy.where(localMin2 <= seedThreshold)
filteredLocalMin = numpy.zeros(localMin.shape, dtype=numpy.uint8)
filteredLocalMin[whereMin] = 1
viewGrayData.append([localMin, "localMin"])
viewGrayData.append([filteredLocalMin, "filteredLocalMin"])
# compute connected components
seeds = vigra.analysis.labelVolumeWithBackground(filteredLocalMin,
neighborhood=26)
viewLabelData.append([seeds, "seeds"])
print "watersheds"
seg, nseg = vigra.analysis.watersheds(pmapG.astype(numpy.float32),
seeds=seeds.astype(numpy.uint32))
print "nseg", nseg
viewLabelData.append([seg, "seg"])
addHocViewer(viewGrayData, viewLabelData, visu=visu)
return se
def addSample(self, f1, f2, label, pluginManager):
# if self.labels == []:
self.labels.append(label)
# else:
# self.labels = np.concatenate((np.array(self.labels),label)) # for adding batches of features
features = self.constructSampleFeatureVector(f1, f2, pluginManager)
if self._numSamples is None:
# use vstack
if self.mydata is None:
self.mydata = features
else:
self.mydata = np.vstack((self.mydata, features))
else:
# allocate full array once, then fill in row by row
if self.mydata is None:
self.mydata = np.zeros((self._numSamples, features.shape[0]))
assert(self._nextIdx < self._numSamples)
self.mydata[self._nextIdx, :] = features
self._nextIdx += 1
def addSample(self, f1, f2, label, pluginManager):
# if self.labels == []:
self.labels.append(label)
# else:
# self.labels = np.concatenate((np.array(self.labels),label)) # for adding batches of features
features = self.constructSampleFeatureVector(f1, f2, pluginManager)
if self._numSamples is None:
# use vstack
if self.mydata is None:
self.mydata = features
else:
self.mydata = np.vstack((self.mydata, features))
else:
# allocate full array once, then fill in row by row
if self.mydata is None:
self.mydata = np.zeros((self._numSamples, features.shape[0]))
assert(self._nextIdx < self._numSamples)
self.mydata[self._nextIdx, :] = features
self._nextIdx += 1
def agglomerative_clustering_2d(img, labels, use_gradient = False): labels = vigra.analysis.labelImage(labels) imgBig = vigra.resize(img, [img.shape[0]*2-1, img.shape[1]*2-1]) if use_gradient: # compute the gradient on interpolated image sigmaGradMag = 2.0 gradMag = vigra.filters.gaussianGradientMagnitude(imgBig, sigmaGradMag) # get 2D grid graph and edgeMap for grid graph # from gradMag of interpolated image or from plain image (depending on use_gradient) gridGraph = vigra.graphs.gridGraph(img.shape[0:2]) gridGraphEdgeIndicator = [] if use_gradient: gridGraphEdgeIndicator = vigra.graphs.edgeFeaturesFromInterpolatedImage(gridGraph, gradMag) else: gridGraphEdgeIndicator = vigra.graphs.edgeFeaturesFromInterpolatedImage(gridGraph, imgBig) # get region adjacency graph from super-pixel labels rag = vigra.graphs.regionAdjacencyGraph(gridGraph, labels) # accumalate edge weights from gradient magintude edgeWeights = rag.accumulateEdgeFeatures(gridGraphEdgeIndicator) # agglomerative clustering beta = 0.5 nodeNumStop = 20 clustering = vigra.graphs.agglomerativeClustering(graph = rag, edgeWeights = edgeWeights, beta = beta, nodeNumStop = nodeNumStop) res = np.zeros( labels.shape ) for c in range(len(clustering)): res[ np.where( labels== (c-1) ) ] = clustering[c] res = vigra.Image(res) res = vigra.analysis.labelImage(res) return res
def computeGlobalFeatures(pmap,
mitoIndex,
membraneIndex,
visu=True,
dsetName='data'):
f = h5py.File(pmap, 'r')
raw = f[dsetName][0:300, 0:300, 300:400, 0]
pMito = f[dsetName][0:300, 0:300, 300:400, membraneIndex].astype('float32')
pMito -= pMito.min()
pMito /= pMito.max()
bMito = numpy.zeros(pMito.shape, dtype='uint8')
bMito[pMito > 0.9] = 1
sdt = vigra.filters.signedDistanceTransform(bMito)
grayData = [(raw, "raw"), (pMito, "pMito"), (bMito, "bMito"), (sdt, "sdt")]
segData = [
#(labels, "labels")
]
skneuro.addHocViewer(grayData, segData, visu=visu)
def computeGlobalFeatures(pmap, mitoIndex, membraneIndex, visu=True, dsetName='data'):
f = h5py.File(pmap,'r')
raw = f[dsetName][0:300,0:300,300:400,0]
pMito = f[dsetName][0:300,0:300,300:400,membraneIndex].astype('float32')
pMito-=pMito.min()
pMito/=pMito.max()
bMito = numpy.zeros(pMito.shape, dtype='uint8')
bMito[pMito>0.9] = 1
sdt = vigra.filters.signedDistanceTransform(bMito)
grayData = [
(raw,"raw"),
(pMito,"pMito"),
(bMito,"bMito"),
(sdt,"sdt")
]
segData = [
#(labels, "labels")
]
skneuro.addHocViewer(grayData, segData,visu=visu)
def getLargeSeeds(raw, pmap, membraneWidth , threshold , rank, visu=False):
with vigra.Timer("ballRankOrderFilter"):
# make mebrane wider with ballRankOrderFilter
r = int(membraneWidth*0.35 + 0.5)
r = max(r, 1)
widerPmap1 = denoise.ballRankOrderFilter(pmap, radius=r, rank=rank)
widerPmap = denoise.ballRankOrderFilter(widerPmap1, radius=r, rank=rank)
widerPmap1 = None
with vigra.Timer("normalize"):
# renormalize
widerPmap -= widerPmap.min()
widerPmap /= widerPmap.max()
with vigra.Timer("binarize"):
# binarize
binaryPmap = numpy.zeros(widerPmap.shape, dtype='uint8')
binaryPmap[pmap>threshold] = 1
#widerPmap = None # save mem
if visu == False:
widerPmap = None
with vigra.Timer("multiBinaryDilation"):
# morphology
# 1) make membrane wider by r
r = int(membraneWidth*1.2 + 0.5)
r = max(r, 2)
mBinaryPmapA = vigra.filters.multiBinaryDilation(binaryPmap,r)
#binaryPmap = None # save mem
if visu == False:
binaryPmap = None
#with vigra.Timer("multiBinaryErosion"):
# # morphology
# # 1) make membrane smaller by r
# r = int(membraneWidth*0.1 + 0.5)
# r = max(r, 1)
# mBinaryPmapB = vigra.filters.multiBinaryErosion(mBinaryPmapA,r)
# if visu == False:
# mBinaryPmapA = None
with vigra.Timer("labelVolumeWithBackground"):
# get seeds
invertedBinaryPmap = 1- mBinaryPmapA
if visu == False:
mBinaryPmapB = None
invertedBinaryPmap = numpy.require(invertedBinaryPmap, dtype='uint32')
ccImg = vigra.analysis.labelVolumeWithBackground(invertedBinaryPmap)
if visu == False:
invertedBinaryPmap = None
if visu:
grayData = [
(raw, "raw"),
(pmap,"pmap"),
(widerPmap,"widerPmap"),
(binaryPmap,"binaryPmap"),
(mBinaryPmapA,"mBinaryPmapA"),
#(mBinaryPmapB,"mBinaryPmapB"),
]
segData = [
(ccImg, "seeds"),
#(seg, "seg")
]
skneuro.addHocViewer(grayData, segData)
return ccImg
# add all internal links of tracklets
for v in uuidToTraxelMap.values():
prev = None
for timestepIdTuple in v:
if prev is not None:
links.append((prev, timestepIdTuple))
prev = timestepIdTuple
# group by timestep
timesteps = [t for t in traxelIdPerTimestepToUniqueIdMap.keys()]
linksPerTimestep = dict([(t, [(a[1], b[1]) for a, b in links
if b[0] == int(t)]) for t in timesteps])
assert (len(linksPerTimestep['0']) == 0)
# create output array
resultVolume = np.zeros((len(timesteps), ) + shape, dtype='uint32')
print("resulting volume shape: {}".format(resultVolume.shape))
progressBar = ProgressBar(stop=len(timesteps))
progressBar.show(0)
# iterate over timesteps and label tracks from front to back in a distinct color
nextUnusedColor = 1
lastFrameColorMap = {}
lastFrameLabelImage = getLabelImageForFrame(args.labelImageFilename,
args.labelImagePath, 0, shape)
for t in range(1, len(timesteps)):
progressBar.show()
thisFrameColorMap = {}
thisFrameLabelImage = getLabelImageForFrame(args.labelImageFilename,
args.labelImagePath, t,
shape)
def getLargeSeeds(raw, pmap, membraneWidth, threshold, rank, visu=False):
with vigra.Timer("ballRankOrderFilter"):
# make mebrane wider with ballRankOrderFilter
r = int(membraneWidth * 0.35 + 0.5)
r = max(r, 1)
widerPmap1 = denoise.ballRankOrderFilter(pmap, radius=r, rank=rank)
widerPmap = denoise.ballRankOrderFilter(widerPmap1,
radius=r,
rank=rank)
widerPmap1 = None
with vigra.Timer("normalize"):
# renormalize
widerPmap -= widerPmap.min()
widerPmap /= widerPmap.max()
with vigra.Timer("binarize"):
# binarize
binaryPmap = numpy.zeros(widerPmap.shape, dtype='uint8')
binaryPmap[pmap > threshold] = 1
#widerPmap = None # save mem
if visu == False:
widerPmap = None
with vigra.Timer("multiBinaryDilation"):
# morphology
# 1) make membrane wider by r
r = int(membraneWidth * 1.2 + 0.5)
r = max(r, 2)
mBinaryPmapA = vigra.filters.multiBinaryDilation(binaryPmap, r)
#binaryPmap = None # save mem
if visu == False:
binaryPmap = None
#with vigra.Timer("multiBinaryErosion"):
# # morphology
# # 1) make membrane smaller by r
# r = int(membraneWidth*0.1 + 0.5)
# r = max(r, 1)
# mBinaryPmapB = vigra.filters.multiBinaryErosion(mBinaryPmapA,r)
# if visu == False:
# mBinaryPmapA = None
with vigra.Timer("labelVolumeWithBackground"):
# get seeds
invertedBinaryPmap = 1 - mBinaryPmapA
if visu == False:
mBinaryPmapB = None
invertedBinaryPmap = numpy.require(invertedBinaryPmap, dtype='uint32')
ccImg = vigra.analysis.labelVolumeWithBackground(invertedBinaryPmap)
if visu == False:
invertedBinaryPmap = None
if visu:
grayData = [
(raw, "raw"),
(pmap, "pmap"),
(widerPmap, "widerPmap"),
(binaryPmap, "binaryPmap"),
(mBinaryPmapA, "mBinaryPmapA"),
#(mBinaryPmapB,"mBinaryPmapB"),
]
segData = [
(ccImg, "seeds"),
#(seg, "seg")
]
skneuro.addHocViewer(grayData, segData)
return ccImg
for v in uuidToTraxelMap.values():
prev = None
for timestepIdTuple in v:
if prev is not None:
links.append((prev, timestepIdTuple))
prev = timestepIdTuple
# group by timestep
# timesteps = [t for t in traxelIdPerTimestepToUniqueIdMap.keys()]
# there might be empty frames. We want them as output too.
timesteps = [str(t).decode("utf-8") for t in range(int(min(traxelIdPerTimestepToUniqueIdMap.keys())) , int(max(traxelIdPerTimestepToUniqueIdMap.keys()))+1 )]
linksPerTimestep = dict([(t, [(a[1], b[1]) for a, b in links if b[0] == int(t)]) for t in timesteps])
assert(len(linksPerTimestep['0']) == 0)
# create output array
resultVolume = np.zeros((len(timesteps),) + shape, dtype='uint32')
print("resulting volume shape: {}".format(resultVolume.shape))
progressBar = ProgressBar(stop=len(timesteps))
progressBar.show(0)
# iterate over timesteps and label tracks from front to back in a distinct color
nextUnusedColor = 1
lastFrameColorMap = {}
lastFrameLabelImage = getLabelImageForFrame(args.labelImageFilename, args.labelImagePath, 0, shape)
for t in range(1,len(timesteps)):
progressBar.show()
thisFrameColorMap = {}
thisFrameLabelImage = getLabelImageForFrame(args.labelImageFilename, args.labelImagePath, t, shape)
for a, b in linksPerTimestep[str(t)]:
# propagate color if possible, otherwise assign a new one
if a in lastFrameColorMap:
if args.overwrite_channel is not None:
threshold_channel = args.overwrite_channel
print("Using user-provided channel: {}".format(threshold_channel))
else:
print("Using channel from project-file: {}".format(threshold_channel))
print("Found PredictionMaps of shape {} (txyzc), using channel {}".format(
predictionMaps.shape, threshold_channel))
progressBar = ProgressBar(stop=shape[0])
progressBar.show(0)
with h5py.File(args.out, 'w') as h5file:
if args.single_volume:
# loop over timesteps
fullLabelImage = np.zeros(
[shape[0], shape[3], shape[2], shape[1], 1], dtype='uint32')
for t in range(shape[0]):
# smooth, threshold, and size filter from prediction map
framePrediction = predictionMaps[t, ...,
threshold_channel].squeeze()
assert ndim == len(framePrediction.shape)
smoothedFramePrediction = vigra.filters.gaussianSmoothing(
framePrediction.astype('float32'), threshold_sigmas[:ndim])
foreground = np.zeros(smoothedFramePrediction.shape,
dtype='uint32')
foreground[smoothedFramePrediction > threshold_level] = 1
if ndim == 2:
labelImage = vigra.analysis.labelImageWithBackground(
foreground)
threshold_channel = args.overwrite_channel
print("Using user-provided channel: {}".format(threshold_channel))
else:
print("Using channel from project-file: {}".format(threshold_channel))
print("Found PredictionMaps of shape {} (txyzc), using channel {}".format(
predictionMaps.shape,
threshold_channel))
progressBar = ProgressBar(stop=shape[0])
progressBar.show(0)
with h5py.File(args.out, 'w') as h5file:
if args.single_volume:
# loop over timesteps
fullLabelImage = np.zeros([shape[0], shape[3], shape[2], shape[1], 1], dtype='uint32')
for t in range(shape[0]):
# smooth, threshold, and size filter from prediction map
framePrediction = predictionMaps[t, ..., threshold_channel].squeeze()
assert ndim == len(framePrediction.shape)
smoothedFramePrediction = vigra.filters.gaussianSmoothing(framePrediction.astype('float32'), threshold_sigmas[:ndim])
foreground = np.zeros(smoothedFramePrediction.shape, dtype='uint32')
foreground[smoothedFramePrediction > threshold_level] = 1
if ndim == 2:
labelImage = vigra.analysis.labelImageWithBackground(foreground)
else:
labelImage = vigra.analysis.labelVolumeWithBackground(foreground)
# filter too small / too large objects out
