def findPeaks(img4D, params):
"""
img4D: a 4D RandomAccessibleInterval
params["frames"]: the number of consecutive time points to average
towards detecting peaks with difference of Gaussian.
Returns a list of lists of peaks found, one list per time point.
"""
frames = params["frames"]
# Work image: the current sum
sum3D = ArrayImgs.unsignedLongs([img4D.dimension(d) for d in [0, 1, 2]])
peaks = []
# Sum of the first set of frames
compute(add([Views.hyperSlice(img4D, 3, i) for i in xrange(frames)])).into(sum3D)
# Extract nuclei from first sum3D
peaks.append(doGPeaks(sum3D, params))
# Running sums: subtract the first and add the last
for i in xrange(frames, img4D.dimension(3), 1):
compute(add(sub(sum3D,
Views.hyperSlice(img4D, 3, i - frames)),
Views.hyperSlice(img4D, 3, i))) \
.into(sum3D)
# Extract nuclei from sum4D
peaks.append(doGPeaks(sum3D, params))
return peaks
def as2DKernel(imgE, weights):
""" imgE: a RandomAccessible, such as an extended view of a RandomAccessibleInterval.
weights: an odd-length list defining a square convolution kernel
centered on the pixel, with columns moving slower than rows.
Returns an ImgMath op.
"""
# Check preconditions: validate kernel
if 1 != len(weights) % 2:
raise Error("list of kernel weights must have an odd length.")
side = int(pow(len(weights), 0.5)) # sqrt
if pow(side, 2) != len(weights):
raise Error("kernel must be a square.")
half = side / 2
# Generate ImgMath ops
# Note that multiplications by weights of value 1 or 0 will be erased automatically
# so that the hierarchy of operations will be the same as in the manual approach above.
return add([mul(weight, offset(imgE, [index % side - half, index / side - half]))
for index, weight in enumerate(weights) if 0 != weight])
c.fwd()
c.localize(pos)
print "Cursor:", pos, "::", c.get()
# Test RandomAccess
ra = iraf.randomAccess()
c = iraf.cursor()
while c.hasNext():
c.fwd()
ra.setPosition(c)
c.localize(pos)
print "RandomAccess:", pos, "::", ra.get()
# Test source img: should be untouched
c = img.cursor()
while c.hasNext():
print "source:", c.next()
# Test interval view: the middle 2x2 square
v = Views.interval(iraf, [1, 1], [2, 2])
IL.wrap(v, "+2 view").show()
# An array from 0 to 15
a = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]
pixels = array(a, 'b')
img = ArrayImgs.unsignedBytes(pixels, [4, 4])
test(add(img, 2).view())
test(add(img, 2).view(FloatType()))
blockHB = block(imgE, cornersHB) op3 = sub(blockHT, blockHB) op4 = sub(blockHB, blockHT) # Two bright-black-bright vertical features 4x8 - 4x8 - 4x8 corners3VL = [[x - 2, y] for x, y in cornersVL] corners3VC = [[x + 2, y] for x, y in cornersVL] corners3VR = [[x + 2, y] for x, y in cornersVR] print corners3VL print corners3VC print corners3VR block3VL = block(imgE, corners3VC) block3VC = block(imgE, corners3VL) block3VR = block(imgE, corners3VR) op5 = sub(block3VC, block3VL, block3VR) # center minus sides op6 = sub(add(block3VL, block3VR), block3VC) # sides minus center # Two bright-black-bright horizontal features 4x8 / 4x8 / 4x8 corners3HT = [[x, y - 2] for x, y in cornersHT] corners3HC = [[x, y + 2] for x, y in cornersHT] corners3HB = [[x, y + 2] for x, y in cornersHB] print corners3HT print corners3HC print corners3HB block3HT = block(imgE, corners3HT) block3HC = block(imgE, corners3HC) block3HB = block(imgE, corners3HB) op7 = sub(block3HC, block3HT, block3HB) # center minus top and bottom op8 = sub(add(block3HT, block3HB), block3HC) # top and bottom minus center for name, op in ((name, eval(name)) for name in vars()
from net.imglib2.type.numeric.real import FloatType ft = FloatType(10) ft.pow(2) print ft from net.imglib2.algorithm.math.ImgMath import compute, add, power from net.imglib2.img.array import ArrayImgs from net.imglib2.img.display.imagej import ImageJFunctions as IL img = ArrayImgs.floats([10, 10, 10]) compute(add(img, 5)).into(img) # in place compute(power(img, 2)).into(img) # in place print 25 * 10 * 10 * 10 == sum(t.get() for t in img.cursor()) IL.wrap(img, "5 squared").show()
def filterBank(img, bs=4, bl=8, sumType=UnsignedLongType(), converter=Util.genericRealTypeConverter()):
""" Haar-like features from Viola and Jones using integral images
tuned to identify neuron membranes in electron microscopy.
bs: length of the short side of a block
bl: length of the long side of a block
sumType: the type with which to add up pixel values in the integral image
converter: for the IntegralImg, to convert from input to sumType
"""
# Create the integral image, stored as 64-bit
alg = IntegralImg(img, sumType, converter)
alg.process()
integralImg = alg.getResult()
imgE = Views.extendBorder(integralImg)
# corners of a 4x8 or 8x4 rectangular block where 0,0 is the top left
cornersV = [[0, 0], [bs -1, 0], # Vertical
[0, bl -1], [bs -1, bl - 1]]
cornersH = [[0, 0], [bl -1, 0], # Horizontal
[0, bs -1], [bl -1, bs - 1]]
# Two adjacent vertical rectangles 4x8 - 4x8 centered on the pixel
blockVL = block(imgE, shift(cornersV, -bs, -bl/2))
blockVR = block(imgE, shift(cornersV, 0, -bl/2))
op1 = let("VL", blockVL,
"VR", blockVR,
IF(GT("VL", "VR"),
THEN(div("VR", "VL")),
ELSE(div("VL", "VR"))))
#op1 = sub(blockVL, blockVR)
#op2 = sub(blockVR, blockVL)
# Two adjacent horizontal rectangles 8x4 - 8x4 centered on the pixel
blockHT = block(imgE, shift(cornersH, -bs, -bl/2))
blockHB = block(imgE, shift(cornersH, -bs, 0))
op3 = let("HT", blockHT,
"HB", blockHB,
IF(GT("HT", "HB"),
THEN(div("HB", "HT")), # div works better than sub
ELSE(div("HT", "HB"))))
#op3 = sub(blockHT, blockHB)
#op4 = sub(blockHB, blockHT)
# Two bright-black-bright vertical features 4x8 - 4x8 - 4x8
block3VL = block(imgE, shift(cornersV, -bs -bs/2, -bl/2))
block3VC = block(imgE, shift(cornersV, -bs/2, -bl/2))
block3VR = block(imgE, shift(cornersV, bs/2, -bl/2))
op5 = let("3VL", block3VL,
"3VC", block3VC,
"3VR", block3VR,
IF(AND(GT("3VL", "3VC"),
GT("3VR", "3VC")),
THEN(sub(add("3VL", "3VR"), "3VC")), # like Viola and Jones 2001
ELSE(div(add("3VL", "3VC", "3VR"), 3)))) # average block value
# Purely like Viola and Jones 2001: work poorly for EM membranes
#op5 = sub(block3VC, block3VL, block3VR) # center minus sides
#op6 = sub(add(block3VL, block3VR), block3VC) # sides minus center
# Two bright-black-bright horizontal features 8x4 / 8x4 / 8x4
block3HT = block(imgE, shift(cornersH, -bl/2, -bs -bs/2))
block3HC = block(imgE, shift(cornersH, -bl/2, -bs/2))
block3HB = block(imgE, shift(cornersH, -bl/2, bs/2))
op7 = let("3HT", block3HT,
"3HC", block3HC,
"3HB", block3HB,
IF(AND(GT("3HT", "3HC"),
GT("3HB", "3HC")),
THEN(sub(add("3HT", "3HB"), "3HC")), # like Viola and Jones 2001
ELSE(div(add("3HT", "3HC", "3HB"), 3)))) # average block value
# Purely like Viola and Jones 2001: work poorly for EM membranes
#op7 = sub(block3HC, block3HT, block3HB) # center minus top and bottom
#op8 = sub(add(block3HT, block3HB), block3HC) # top and bottom minus center
# Combination of vertical and horizontal edge detection
#op9 = maximum(op1, op3)
#op10 = maximum(op6, op8)
#op10 = maximum(op5, op7)
"""
# corners of a square block where 0,0 is at the top left
cornersS = [[0, 0], [bs, 0],
[0, bs], [bs, bs]]
# 2x2 squares for oblique edge detection
blockSTL = block(imgE, shift(cornersS, -bs, -bs)) # top left
blockSTR = block(imgE, shift(cornersS, 0, -bs)) # top right
blockSBL = block(imgE, shift(cornersS, -bs, 0)) # bottom left
blockSBR = block(imgE, cornersS) # bottom right
op11 = sub(add(blockSTL, blockSBR), blockSTR, blockSBL)
op12 = sub(add(blockSTR, blockSBL), blockSTL, blockSBR)
"""
# Combination of vertical, horizontal and oblique edge detection
#op13 = maximum(op1, op3, op6, op8, op11, op12)
#op13 = maximum(op1, op3, op5, op7, op11, op12) # combinations are terrible for RandomForest
# Edge detectors: sum of 3 adjacent pixels (not dividing by the other 6
# to avoid penalizing Y membrane configurations)
"""
op14 = maximum(add(offset(op13, [-1, -1]), op13, offset(op13, [ 1, 1])),
add(offset(op13, [ 0, -1]), op13, offset(op13, [ 0, 1])),
add(offset(op13, [ 1, -1]), op13, offset(op13, [-1, 1])),
add(offset(op13, [-1, 0]), op13, offset(op13, [ 1, 0])))
"""
#opMean, opVariance = filterBankBlockStatistics(img, integralImgE=imgE, sumType=sumType, converter=converter)
# Return an ordered list of all ops
#return [op1, op2, op3, op4, op5, op6, op7, op8, op9, op10, op11, op12, op13, op14]
#return [op1, op3, op5, op7, opMean, opVariance]
return [op1, op3, op5, op7]
def filterBankOrthogonalEdges(img,
bs=4,
bl=8,
sumType=UnsignedLongType(),
converter=Util.genericRealTypeConverter()):
""" Haar-like features from Viola and Jones using integral images of a set of rotated images
tuned to identify neuron membranes in electron microscopy.
bs: length of the short side of a block
bl: length of the long side of a block
sumType: the type with which to add up pixel values in the integral image
converter: for the IntegralImg, to convert from input to sumType
"""
# Create the integral image, stored as 64-bit
alg = IntegralImg(img, sumType, converter)
alg.process()
integralImg = alg.getResult()
imgE = Views.extendBorder(integralImg)
# corners of a 4x8 or 8x4 rectangular block where 0,0 is the top left
cornersV = [[0, 0], [bs -1, 0], # Vertical
[0, bl -1], [bs -1, bl - 1]]
cornersH = [[0, 0], [bl -1, 0], # Horizontal
[0, bs -1], [bl -1, bs - 1]]
# Two adjacent vertical rectangles 4x8 - 4x8 centered on the pixel
blockVL = block(imgE, shift(cornersV, -bs, -bl/2))
blockVR = block(imgE, shift(cornersV, 0, -bl/2))
op1 = sub(blockVL, blockVR)
op2 = sub(blockVR, blockVL)
# Two adjacent horizontal rectangles 8x4 - 8x4 centered on the pixel
blockHT = block(imgE, shift(cornersH, -bs, -bl/2))
blockHB = block(imgE, shift(cornersH, -bs, 0))
op3 = sub(blockHT, blockHB)
op4 = sub(blockHB, blockHT)
# Two bright-black-bright vertical features 4x8 - 4x8 - 4x8
block3VL = block(imgE, shift(cornersV, -bs -bs/2, -bl/2))
block3VC = block(imgE, shift(cornersV, -bs/2, -bl/2))
block3VR = block(imgE, shift(cornersV, bs/2, -bl/2))
op5 = let("3VL", block3VL,
"3VC", block3VC,
"3VR", block3VR,
IF(AND(GT("3VL", "3VC"),
GT("3VR", "3VC")),
THEN(sub(add("3VL", "3VR"), "3VC")), # like Viola and Jones 2001
ELSE(div(add("3VL", "3VC", "3VR"), 3)))) # average block value
# Purely like Viola and Jones 2001: work poorly for EM membranes
#op5 = sub(block3VC, block3VL, block3VR) # center minus sides
#op6 = sub(add(block3VL, block3VR), block3VC) # sides minus center
# Two bright-black-bright horizontal features 8x4 / 8x4 / 8x4
block3HT = block(imgE, shift(cornersH, -bl/2, -bs -bs/2))
block3HC = block(imgE, shift(cornersH, -bl/2, -bs/2))
block3HB = block(imgE, shift(cornersH, -bl/2, bs/2))
op7 = let("3HT", block3HT,
"3HC", block3HC,
"3HB", block3HB,
IF(AND(GT("3HT", "3HC"),
GT("3HB", "3HC")),
THEN(sub(add("3HT", "3HB"), "3HC")),
ELSE(div(add("3HT", "3HC", "3HB"), 3)))) # average block value TODO make a single block
#op7 = sub(block3HC, block3HT, block3HB) # center minus top and bottom
#op8 = sub(add(block3HT, block3HB), block3HC) # top and bottom minus center
# Two adjacent horizontal rectanges, 12x12 - 4x4
bll = bl + bl/2
cornersLarge = [[0, 0], [bll -1, 0],
[0, bll -1], [bll -1, bll -1]]
cornersSmall = [[0, 0], [bs -1, 0],
[0, bs -1], [bs -1, bs -1]]
# Subtract a large black rectangle from a small bright one - aiming at capturing synapses
# Bright on the right
blockLargeL = block(imgE, shift(cornersLarge, -bll, -bll/2))
blockSmallR = block(imgE, shift(cornersSmall, 0, -bs/2))
op9 = sub(blockSmallR, blockLargeL)
# Bright on the left
blockLargeR = block(imgE, shift(cornersLarge, 0, -bll/2))
blockSmallL = block(imgE, shift(cornersSmall, -bs, -bs/2))
op10 = sub(blockSmallL, blockLargeR)
# Bright at the bottom
blockLargeT = block(imgE, shift(cornersLarge, -bll/2, -bll))
blockSmallB = block(imgE, shift(cornersSmall, -bs/2, 0))
op11 = sub(blockSmallB, blockLargeT)
# Bright at the top
blockLargeB = block(imgE, shift(cornersLarge, -bll/2, 0))
blockSmallT = block(imgE, shift(cornersSmall, -bs/2, -bs))
op12 = sub(blockSmallT, blockLargeB)
return [op1, op2, op3, op4, op5, op7, op9, op10, op11, op12]
# Create the integral image of an 8-bit input, stored as 64-bit
target = ArrayImgs.unsignedLongs(Intervals.dimensionsAsLongArray(img))
# Copy input onto the target image
compute(img).into(target)
# Extend target with zeros, so that we can read at coordinate -1
imgE = Views.extendZero(target)
# Integrate every dimension, cummulatively by writing into
# a target image that is also the input
for d in xrange(img.numDimensions()):
coord = [0] * img.numDimensions() # array of zeros
coord[d] = -1
# Cummulative sum along the current dimension
# Note that instead of the ImgMath offset op,
# we could have used Views.translate(Views.extendZero(target), [1, 0]))
# (Notice though the sign change in the translation)
integral = add(target, offset(imgE, coord))
compute(integral).into(target)
# The target is the integral image
integralImg = target
# Read out blocks of radius 5 (i.e. 10x10 for a 2d image)
# in a way that is entirely n-dimensional (applies to 1d, 2d, 3d, 4d ...)
radius = 5
nd = img.numDimensions()
op = div(block(Views.extendBorder(integralImg), [radius] * nd),
pow(radius * 2, nd)) # divide by total number of pixels in the block
blurred = ArrayImgs.floats(Intervals.dimensionsAsLongArray(img))
compute(op).into(blurred, FloatType())
blockVR = block(imgE, shift(cornersV, 0, -bl / 2)) op1 = sub(blockVL, blockVR) op2 = sub(blockVR, blockVL) # Two adjacent horizontal rectangles 8x4 - 8x4 centered on the pixel blockHT = block(imgE, shift(cornersH, -bs, -bl / 2)) blockHB = block(imgE, shift(cornersH, -bs, 0)) op3 = sub(blockHT, blockHB) op4 = sub(blockHB, blockHT) # Two bright-black-bright vertical features 4x8 - 4x8 - 4x8 block3VL = block(imgE, shift(cornersV, -bs - bs / 2, -bl / 2)) block3VC = block(imgE, shift(cornersV, -bs / 2, -bl / 2)) block3VR = block(imgE, shift(cornersV, bs / 2, -bl / 2)) op5 = sub(block3VC, block3VL, block3VR) # center minus sides op6 = sub(add(block3VL, block3VR), block3VC) # sides minus center # Two bright-black-bright horizontal features 8x4 / 8x4 / 8x4 block3HT = block(imgE, shift(cornersH, -bl / 2, -bs - bs / 2)) block3HC = block(imgE, shift(cornersH, -bl / 2, -bs / 2)) block3HB = block(imgE, shift(cornersH, -bl / 2, bs / 2)) op7 = sub(block3HC, block3HT, block3HB) # center minus top and bottom op8 = sub(add(block3HT, block3HB), block3HC) # top and bottom minus center # Combination of vertical and horizontal edge detection op9 = maximum(op1, op3) op10 = maximum(op6, op8) # corners of a square block where 0,0 is at the top left cornersS = [[0, 0], [bs, 0], [0, bs], [bs, bs]]
def filterBank(img,
sumType=UnsignedLongType(),
converter=Util.genericRealTypeConverter()):
""" Haar-like features from Viola and Jones
tuned to identify neuron membranes in electron microscopy. """
# Create the integral image, stored as 64-bit
alg = IntegralImg(img, sumType, converter)
alg.process()
integralImg = alg.getResult()
imgE = Views.extendBorder(integralImg)
# corners of a 4x8 or 8x4 rectangular block where 0,0 is the top left
bs = 4 # short side
bl = 8 # long side
cornersV = [
[0, 0],
[bs - 1, 0], # Vertical
[0, bl - 1],
[bs - 1, bl - 1]
]
cornersH = [
[0, 0],
[bl - 1, 0], # Horizontal
[0, bs - 1],
[bl - 1, bs - 1]
]
# Two adjacent vertical rectangles 4x8 - 4x8 centered on the pixel
blockVL = block(imgE, shift(cornersV, -bs, -bl / 2))
blockVR = block(imgE, shift(cornersV, 0, -bl / 2))
op1 = sub(blockVL, blockVR)
op2 = sub(blockVR, blockVL)
# Two adjacent horizontal rectangles 8x4 - 8x4 centered on the pixel
blockHT = block(imgE, shift(cornersH, -bs, -bl / 2))
blockHB = block(imgE, shift(cornersH, -bs, 0))
op3 = sub(blockHT, blockHB)
op4 = sub(blockHB, blockHT)
# Two bright-black-bright vertical features 4x8 - 4x8 - 4x8
block3VL = block(imgE, shift(cornersV, -bs - bs / 2, -bl / 2))
block3VC = block(imgE, shift(cornersV, -bs / 2, -bl / 2))
block3VR = block(imgE, shift(cornersV, bs / 2, -bl / 2))
op5 = sub(block3VC, block3VL, block3VR) # center minus sides
op6 = sub(add(block3VL, block3VR), block3VC) # sides minus center
# Two bright-black-bright horizontal features 8x4 / 8x4 / 8x4
block3HT = block(imgE, shift(cornersH, -bl / 2, -bs - bs / 2))
block3HC = block(imgE, shift(cornersH, -bl / 2, -bs / 2))
block3HB = block(imgE, shift(cornersH, -bl / 2, bs / 2))
op7 = sub(block3HC, block3HT, block3HB) # center minus top and bottom
op8 = sub(add(block3HT, block3HB), block3HC) # top and bottom minus center
# Combination of vertical and horizontal edge detection
op9 = maximum(op1, op3)
op10 = maximum(op6, op8)
# corners of a square block where 0,0 is at the top left
cornersS = [[0, 0], [bs, 0], [0, bs], [bs, bs]]
# 2x2 squares for oblique edge detection
blockSTL = block(imgE, shift(cornersS, -bs, -bs)) # top left
blockSTR = block(imgE, shift(cornersS, 0, -bs)) # top right
blockSBL = block(imgE, shift(cornersS, -bs, 0)) # bottom left
blockSBR = block(imgE, cornersS) # bottom right
op11 = sub(add(blockSTL, blockSBR), blockSTR, blockSBL)
op12 = sub(add(blockSTR, blockSBL), blockSTL, blockSBR)
# Combination of vertical, horizontal and oblique edge detection
op13 = maximum(op1, op3, op6, op8, op11, op12)
# Edge detectors: sum of 3 adjacent pixels (not dividing by the other 6
# to avoid penalizing Y membrane configurations)
op14 = maximum(add(offset(op13, [-1, -1]), op13, offset(op13, [1, 1])),
add(offset(op13, [0, -1]), op13, offset(op13, [0, 1])),
add(offset(op13, [1, -1]), op13, offset(op13, [-1, 1])),
add(offset(op13, [-1, 0]), op13, offset(op13, [1, 0])))
# Return a list of all ops
#return [ob for name, ob in vars().iteritems() if re.match(r"^op\d+$", name)]
# Ordered
return [
op1, op2, op3, op4, op5, op6, op7, op8, op9, op10, op11, op12, op13,
op14
]
def test(red, green, blue, easy=True):
saturation = let(
"red", red, "green", green, "blue", blue, "max",
maximum("red", "green", "blue"), "min",
minimum("red", "green", "blue"),
IF(EQ(0, "max"), THEN(0), ELSE(div(sub("max", "min"), "max"))))
brightness = div(maximum(red, green, blue), 255.0)
hue = IF(
EQ(0, saturation), THEN(0),
ELSE(
let(
"red", red, "green", green, "blue", blue, "max",
maximum("red", "green", "blue"), "min",
minimum("red", "green", "blue"), "range", sub("max", "min"),
"redc", div(sub("max", "red"), "range"), "greenc",
div(sub("max", "green"), "range"), "bluec",
div(sub("max", "blue"), "range"), "hue",
div(
IF(
EQ("red", "max"), THEN(sub("bluec", "greenc")),
ELSE(
IF(EQ("green", "max"),
THEN(sub(add(2, "redc"), "bluec")),
ELSE(sub(add(4, "greenc"), "redc"))))), 6),
IF(LT("hue", 0), THEN(add("hue", 1)), ELSE("hue")))))
#print hierarchy(hue)
#print "hue view:", hue.view( FloatType() ).iterationOrder()
if easy:
# About 26 ms
"""
hsb = Views.stack( hue.view( FloatType() ),
saturation.view( FloatType() ),
brightness.view( FloatType() ) )
"""
# About 13 ms: half! Still much worse than plain ImageJ,
# but the source images are iterated 4 times, rather than just once,
# and the saturation is computed twice,
# and the min, max is computed 3 and 4 times, respectively.
hsb = Views.stack(hue.viewDouble(FloatType()),
saturation.viewDouble(FloatType()),
brightness.viewDouble(FloatType()))
"""
# Even worse: ~37 ms
width, height = rgb.dimension(0), rgb.dimension(1)
h = compute(hue).into(ArrayImgs.floats([width, height]))
s = compute(saturation).into(ArrayImgs.floats([width, height]))
b = compute(brightness).into(ArrayImgs.floats([width, height]))
hsb = Views.stack( h, s, b )
"""
imp = IL.wrap(hsb, "HSB view")
else:
# Tested it: takes more time (~40 ms vs 26 ms above)
width, height = rgb.dimension(0), rgb.dimension(1)
hb = zeros(width * height, 'f')
sb = zeros(width * height, 'f')
bb = zeros(width * height, 'f')
h = ArrayImgs.floats(hb, [width, height])
s = ArrayImgs.floats(sb, [width, height])
b = ArrayImgs.floats(bb, [width, height])
#print "ArrayImg:", b.iterationOrder()
ImgUtil.copy(ImgView.wrap(hue.view(FloatType()), None), h)
ImgUtil.copy(ImgView.wrap(saturation.view(FloatType()), None), s)
ImgUtil.copy(ImgView.wrap(brightness.view(FloatType()), None), b)
stack = ImageStack(width, height)
stack.addSlice(FloatProcessor(width, height, hb, None))
stack.addSlice(FloatProcessor(width, height, sb, None))
stack.addSlice(FloatProcessor(width, height, bb, None))
imp = ImagePlus("hsb", stack)
return imp