def test_03_04_quartiles(self):
"""test quartile values on a 250x250 square filled with uniform values"""
labels = np.ones((250, 250), int)
np.random.seed(0)
image = np.random.uniform(size=(250, 250))
ii = II.InjectImage('MyImage', image)
ii.module_num = 1
io = II.InjectObjects('MyObjects', labels)
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
feature_name = '%s_%s_%s' % (MOI.INTENSITY,
MOI.LOWER_QUARTILE_INTENSITY, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], .25, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY, MOI.MEDIAN_INTENSITY,
'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], .50, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY, MOI.MAD_INTENSITY,
'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], 0.25, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY,
MOI.UPPER_QUARTILE_INTENSITY, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], .75, 2)
def test_03_01_01_masked(self):
"""Make sure we can process a completely masked image
Regression test of IMG-971
"""
ii = II.InjectImage('MyImage', np.zeros((10, 10)),
np.zeros((10, 10), bool))
ii.module_num = 1
io = II.InjectObjects('MyObjects', np.ones((10, 10), int))
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
for meas_name in MOI.ALL_MEASUREMENTS:
feature_name = "%s_%s_%s" % (MOI.INTENSITY, meas_name, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertEqual(np.product(data.shape), 1)
self.assertTrue(np.all(np.isnan(data) | (data == 0)))
self.features_and_columns_match(m, moi)
def test_03_01_00_zero(self):
"""Make sure we can process a blank image"""
ii = II.InjectImage('MyImage', np.zeros((10, 10)))
ii.module_num = 1
io = II.InjectObjects('MyObjects', np.zeros((10, 10), int))
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
for category, features in ((MOI.INTENSITY, MOI.ALL_MEASUREMENTS),
(MOI.C_LOCATION,
MOI.ALL_LOCATION_MEASUREMENTS)):
for meas_name in features:
feature_name = "%s_%s_%s" % (category, meas_name, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertEqual(np.product(data.shape), 0,
"Got data for feature %s" % feature_name)
self.features_and_columns_match(m, moi)
def test_01_01_combine(self):
img = self.get_my_image()
inj = cpm_inject.InjectImage("my_image", img)
inj.module_num = 1
ctg = cpm_ctg.ColorToGray()
ctg.module_num = 2
ctg.image_name.value = "my_image"
ctg.combine_or_split.value = cpm_ctg.COMBINE
ctg.red_contribution.value = 1
ctg.green_contribution.value = 2
ctg.blue_contribution.value = 3
ctg.grayscale_name.value = "my_grayscale"
pipeline = cpp.Pipeline()
pipeline.add_module(inj)
pipeline.add_module(ctg)
pipeline.test_valid()
measurements = cpm.Measurements()
object_set = cpo.ObjectSet()
image_set_list = cpi.ImageSetList()
workspace = Workspace(pipeline, inj, None, None, measurements,
image_set_list, None)
inj.prepare_run(workspace)
inj.prepare_group(workspace, {}, [1])
image_set = image_set_list.get_image_set(0)
inj.run(Workspace(pipeline, inj, image_set, object_set, measurements, None))
ctg.run(Workspace(pipeline, ctg, image_set, object_set, measurements, None))
grayscale = image_set.get_image("my_grayscale")
self.assertTrue(grayscale)
img = grayscale.image
self.assertAlmostEqual(img[0, 0], 1.0 / 6.0)
self.assertAlmostEqual(img[0, 25], 1.0 / 3.0)
self.assertAlmostEqual(img[25, 0], 1.0 / 2.0)
self.assertAlmostEqual(img[25, 25], 0)
def test_02_02_color_by_color(self):
'''Correct a color image with a black & white illumination fn'''
np.random.seed(0)
image = np.random.uniform(size=(10,10,3)).astype(np.float32)
illum = np.random.uniform(size=(10,10,3)).astype(np.float32)
expected = image - illum
expected[expected < 0] = 0
pipeline = cpp.Pipeline()
pipeline.add_listener(self.error_callback)
input_module = inj.InjectImage("InputImage",image)
input_module.module_num = 1
pipeline.add_module(input_module)
illum_module = inj.InjectImage("IllumImage", illum)
illum_module.module_num = 2
pipeline.add_module(illum_module)
module = cpmcia.CorrectIlluminationApply()
module.module_num = 3
pipeline.add_module(module)
image = module.images[0]
image.image_name.value = "InputImage"
image.illum_correct_function_image_name.value = "IllumImage"
image.corrected_image_name.value = "OutputImage"
image.divide_or_subtract.value = cpmcia.DOS_SUBTRACT
image.rescale_option = cpmcia.RE_NONE
measurements = cpm.Measurements()
image_set_list = cpi.ImageSetList()
measurements = cpm.Measurements()
workspace = cpw.Workspace(pipeline, None, None, None,
measurements, image_set_list)
pipeline.prepare_run(workspace)
input_module.prepare_group(workspace, {}, [1])
illum_module.prepare_group(workspace, {}, [1])
module.prepare_group(workspace, {}, [1])
image_set = image_set_list.get_image_set(0)
object_set = cpo.ObjectSet()
workspace = cpw.Workspace(pipeline,
input_module,
image_set,
object_set,
measurements,
image_set_list)
input_module.run(workspace)
illum_module.run(workspace)
module.run(workspace)
output_image = workspace.image_set.get_image("OutputImage")
self.assertTrue(np.all(output_image.pixel_data == expected))
def test_03_03_01_mass_displacement_masked(self):
"""Regression test IMG-766 - mass displacement of a masked image"""
labels = np.array([[0, 0, 0, 0, 0, 0, 0],
[0, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 0],
[0, 1, 1, 1, 1, 1, 0],
[0, 0, 0, 0, 0, 0, 0],
[0, 2, 2, 2, 2, 2, 0],
[0, 2, 2, 2, 2, 2, 0],
[0, 2, 2, 2, 2, 2, 0],
[0, 2, 2, 2, 2, 2, 0],
[0, 2, 2, 2, 2, 2, 0],
[0, 0, 0, 0, 0, 0, 0],
[0, 3, 3, 3, 3, 3, 0],
[0, 3, 3, 3, 3, 3, 0],
[0, 3, 3, 3, 3, 3, 0],
[0, 3, 3, 3, 3, 3, 0],
[0, 3, 3, 3, 3, 3, 0],
[0, 0, 0, 0, 0, 0, 0]])
image = np.zeros(labels.shape, dtype=float)
#
# image # 1 has a single value in one of the corners
# whose distance is sqrt(8) from the center
#
image[1, 1] = 1
# image # 2 has a single value on the top edge
# and should have distance 2
#
image[7, 3] = 1
# image # 3 has a single value on the left edge
# and should have distance 2
image[15, 1] = 1
mask = np.zeros(image.shape, bool)
mask[labels > 0] = True
ii = II.InjectImage('MyImage', image, mask)
ii.module_num = 1
io = II.InjectObjects('MyObjects', labels)
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
feature_name = '%s_%s_%s' % (MOI.INTENSITY, MOI.MASS_DISPLACEMENT, 'MyImage')
mass_displacement = m.get_current_measurement('MyObjects', feature_name)
self.assertEqual(np.product(mass_displacement.shape), 3)
self.assertAlmostEqual(mass_displacement[0], math.sqrt(8.0))
self.assertAlmostEqual(mass_displacement[1], 2.0)
self.assertAlmostEqual(mass_displacement[2], 2.0)
def test_03_06_quartiles(self):
"""test quartile values on a 250x250 square with 4 objects"""
labels = np.ones((250, 250), int)
labels[125:, :] += 1
labels[:, 125:] += 2
np.random.seed(0)
image = np.random.uniform(size=(250, 250))
#
# Make the distributions center around .5, .25, 1/6 and .125
#
image /= labels.astype(float)
ii = II.InjectImage('MyImage', image)
ii.module_num = 1
io = II.InjectObjects('MyObjects', labels)
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
feature_name = '%s_%s_%s' % (MOI.INTENSITY,
MOI.LOWER_QUARTILE_INTENSITY, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], 1.0 / 4.0, 2)
self.assertAlmostEqual(data[1], 1.0 / 8.0, 2)
self.assertAlmostEqual(data[2], 1.0 / 12.0, 2)
self.assertAlmostEqual(data[3], 1.0 / 16.0, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY, MOI.MEDIAN_INTENSITY,
'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], 1.0 / 2.0, 2)
self.assertAlmostEqual(data[1], 1.0 / 4.0, 2)
self.assertAlmostEqual(data[2], 1.0 / 6.0, 2)
self.assertAlmostEqual(data[3], 1.0 / 8.0, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY,
MOI.UPPER_QUARTILE_INTENSITY, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], 3.0 / 4.0, 2)
self.assertAlmostEqual(data[1], 3.0 / 8.0, 2)
self.assertAlmostEqual(data[2], 3.0 / 12.0, 2)
self.assertAlmostEqual(data[3], 3.0 / 16.0, 2)
feature_name = '%s_%s_%s' % (MOI.INTENSITY, MOI.MAD_INTENSITY,
'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertAlmostEqual(data[0], 1.0 / 4.0, 2)
self.assertAlmostEqual(data[1], 1.0 / 8.0, 2)
self.assertAlmostEqual(data[2], 1.0 / 12.0, 2)
self.assertAlmostEqual(data[3], 1.0 / 16.0, 2)
def test_01_02_split_all(self):
img = self.get_my_image()
inj = cpm_inject.InjectImage("my_image", img)
inj.module_num = 1
ctg = cpm_ctg.ColorToGray()
ctg.module_num = 2
ctg.image_name.value = "my_image"
ctg.combine_or_split.value = cpm_ctg.SPLIT
ctg.use_red.value = True
ctg.use_blue.value = True
ctg.use_green.value = True
ctg.red_name.value = "my_red"
ctg.green_name.value = "my_green"
ctg.blue_name.value = "my_blue"
pipeline = cpp.Pipeline()
pipeline.add_module(inj)
pipeline.add_module(ctg)
pipeline.test_valid()
measurements = cpm.Measurements()
object_set = cpo.ObjectSet()
image_set_list = cpi.ImageSetList()
workspace = Workspace(pipeline, inj, None, None, measurements,
image_set_list, None)
inj.prepare_run(workspace)
inj.prepare_group(workspace, {}, [1])
image_set = image_set_list.get_image_set(0)
inj.run(
Workspace(pipeline, inj, image_set, object_set, measurements,
None))
ctg.run(
Workspace(pipeline, ctg, image_set, object_set, measurements,
None))
red = image_set.get_image("my_red")
self.assertTrue(red)
img = red.image
self.assertAlmostEqual(img[0, 0], 1)
self.assertAlmostEqual(img[0, 25], 0)
self.assertAlmostEqual(img[25, 0], 0)
self.assertAlmostEqual(img[25, 25], 0)
green = image_set.get_image("my_green")
self.assertTrue(green)
img = green.image
self.assertAlmostEqual(img[0, 0], 0)
self.assertAlmostEqual(img[0, 25], 1)
self.assertAlmostEqual(img[25, 0], 0)
self.assertAlmostEqual(img[25, 25], 0)
blue = image_set.get_image("my_blue")
self.assertTrue(blue)
img = blue.image
self.assertAlmostEqual(img[0, 0], 0)
self.assertAlmostEqual(img[0, 25], 0)
self.assertAlmostEqual(img[25, 0], 1)
self.assertAlmostEqual(img[25, 25], 0)
def test_03_02_00_one(self):
"""Check measurements on a 3x3 square of 1's"""
img = np.array([[0, 0, 0, 0, 0, 0, 0], [0, 0, 1, 1, 1, 0, 0],
[0, 0, 1, 1, 1, 0, 0], [0, 0, 1, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0]])
ii = II.InjectImage('MyImage', img.astype(float))
ii.module_num = 1
io = II.InjectObjects('MyObjects', img.astype(int))
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
m = pipeline.run()
for category, meas_name, value in (
(MOI.INTENSITY, MOI.INTEGRATED_INTENSITY,
9), (MOI.INTENSITY, MOI.MEAN_INTENSITY,
1), (MOI.INTENSITY, MOI.STD_INTENSITY,
0), (MOI.INTENSITY, MOI.MIN_INTENSITY,
1), (MOI.INTENSITY, MOI.MAX_INTENSITY, 1),
(MOI.INTENSITY, MOI.INTEGRATED_INTENSITY_EDGE,
8), (MOI.INTENSITY, MOI.MEAN_INTENSITY_EDGE,
1), (MOI.INTENSITY, MOI.STD_INTENSITY_EDGE,
0), (MOI.INTENSITY, MOI.MIN_INTENSITY_EDGE,
1), (MOI.INTENSITY, MOI.MAX_INTENSITY_EDGE,
1), (MOI.INTENSITY, MOI.MASS_DISPLACEMENT, 0),
(MOI.INTENSITY, MOI.LOWER_QUARTILE_INTENSITY,
1), (MOI.INTENSITY, MOI.MEDIAN_INTENSITY,
1), (MOI.INTENSITY, MOI.UPPER_QUARTILE_INTENSITY,
1), (MOI.C_LOCATION, MOI.LOC_CMI_X,
3), (MOI.C_LOCATION, MOI.LOC_CMI_Y, 2)):
feature_name = "%s_%s_%s" % (category, meas_name, 'MyImage')
data = m.get_current_measurement('MyObjects', feature_name)
self.assertEqual(np.product(data.shape), 1)
self.assertEqual(
data[0], value,
"%s expected %f != actual %f" % (meas_name, value, data[0]))
def test_03_05_quartiles(self):
"""Regression test a bug that occurs in an image with one pixel"""
labels = np.zeros((10, 20))
labels[2:7, 3:8] = 1
labels[5, 15] = 2
np.random.seed(0)
image = np.random.uniform(size=(10, 20))
ii = II.InjectImage('MyImage', image)
ii.module_num = 1
io = II.InjectObjects('MyObjects', labels)
io.module_num = 2
moi = MOI.MeasureObjectIntensity()
moi.images[0].name.value = 'MyImage'
moi.objects[0].name.value = 'MyObjects'
moi.module_num = 3
pipeline = P.Pipeline()
pipeline.add_listener(self.error_callback)
pipeline.add_module(ii)
pipeline.add_module(io)
pipeline.add_module(moi)
# Crashes when pipeline runs in measureobjectintensity.py revision 7146
m = pipeline.run()
