def low_signal_threshold(container, _id=0):
"""
Creates a grain mask by blurring the image and then doing Otsu\'s power scaled threshold.
The gaussian filtered image is added to the container.
Parameters
----------
container: gwy.container
_id: int
Key of data in container to process.
"""
# get sigma from settings, ie. value last used in toolbox->filters
sigma = get_gaussian_sigma()
# create undo point
gwy.gwy_app_undo_checkpoint(container, container.keys_by_name())
# get datafield 0 for each container for func to operate on
data = container[gwy.gwy_app_get_data_key_for_id(_id)]
# create new blurred data
blurred = data.duplicate()
# do gaussian filter
blurred.filter_gaussian(sigma)
# add to container
gwy.gwy_app_data_browser_add_data_field(blurred, container, False)
# create grain mask
grains = threshold_otsu_power(blurred)
# add grain field to container and set mask colour
container.set_object_by_name(gwy.gwy_app_get_mask_key_for_id(_id), grains)
set_mask_colour(container, _id)
def run():
# do thresholding on main (id=0) data
_id = 0
# get containers
containers = gwy.gwy_app_data_browser_get_containers()
# get current image data
for container in containers:
# get datafield and compute threshold
datafield = container.get_object(gwy.gwy_app_get_data_key_for_id(_id))
# create grain mask
grains = create_mask(datafield, datafield.otsu_threshold())
# add mask field to container
container.set_object(gwy.gwy_app_get_mask_key_for_id(_id), grains)
# set mask colour and opacity
set_mask_colour(container, _id)
def set_mask_colour(container, _id, rgba=(0.75, 0.0, 0.0, 0.5)):
"""
Convenience function to set mask colour parameters within container.
Parameters
----------
container: gwy.Container
_id: int
Mask key in container.
rgba: tuple of floats, length 4
Values range from 0..1. (Red, Green, Blue, Alpha).
"""
mask_key = gwy.gwy_app_get_mask_key_for_id(_id)
# set mask colour and opacity
for color, val in zip(("red", "green", "blue", "alpha"), rgba):
container[os.path.join(mask_key, color)] = val
def run():
# get current data and correct container for file
_id = gwy.gwy_app_data_browser_get_current(gwy.APP_DATA_FIELD_ID)
container = gwy.gwy_app_data_browser_get_current(gwy.APP_CONTAINER)
# create undo point
gwy.gwy_undo_qcheckpoint(container, container.keys())
# get datafield and compute threshold
datafield = container.get_object(gwy.gwy_app_get_data_key_for_id(_id))
threshold = threshold_otsu_median(datafield)
# create grain mask
grains = create_mask(datafield, threshold)
# add mask field to container
container.set_object(gwy.gwy_app_get_mask_key_for_id(_id), grains)
# set mask colour and opacity
set_mask_colour(container, _id)
def analyse_particles_otsu_median(container,
datafield_id=0,
filter_size=5,
save=False):
"""
Compute grains on an image through histogram clustering, with the median of
the class (rather than the mean) as the class descriptor.
(Translates closely to the mode of class z-scale histogram).
DOI: 10.1016/j.aasri.2012.11.074
Calculates threshold, then puts a mask on the image. Grains are saved
as JSON and .gwy files.
Parameters
----------
container: gwy.Container
Container where data is stored/shown.
datafield_id: int
The id of the data to analyse within container. Deafault is 0, ie. first data channel.
filter_size: int
Remove found grains smaller than this, in pixels. Default is 5.
save: boolean
Whether to save the data as JSON and .gwy. Default is True.
Returns
-------
grain_datafield: gwy.DataField
Binary datafield where grains are foreground.
"""
# key for putting data in container
data_key = os.path.join(os.sep, str(datafield_id), "data")
mask_key = os.path.join(os.sep, str(datafield_id), "mask")
# get datafield
datafield = container.get_object_by_name(
gwy.gwy_app_get_data_key_for_id(datafield_id))
# calculate threhshold
threshold = threshold_otsu_median(datafield)
# --- CREATE MASK BELOW AND ADD TO CONTAINER
grain_datafield = datafield.new_alike(True)
# get relative threshold
rel_threshold = 100.0 * get_relative_value(threshold, datafield)
# threshold --> false means mark above threshold
datafield.grains_mark_height(grain_datafield, rel_threshold, False)
# NB. reordered operations remove_touching_border and area_filter
# on date 5.2.2019
# before this date order went border then size
# grain_datafield.grains_remove_by_size(filter_size)
# added 5.2.2019
# grain_datafield_full = grain_datafield.duplicate()
# container.set_object_by_name(os.path.join(os.sep, str(datafield_id), 'mask_including_edges'), \
# grain_datafield_full)
# remove grains touching image edegs -> skews data
# grain_datafield.grains_remove_touching_border()
container.set_object_by_name(gwy.gwy_app_get_mask_key_for_id(datafield_id),
grain_datafield)
# set mask colour and opacity
container[os.path.join(mask_key, "alpha")] = 0.25
container[os.path.join(mask_key, "blue")] = 1.0
if save:
save_dict = dict(
THRESHOLD=threshold,
RELATIVE_THRESHOLD=rel_threshold,
GRAIN_DATAFIELD_INCLUDING_EDGES=grain_datafield_full.get_data(),
)
save_gwy_grain_mask_to_JSON(container,
datafield_id=datafield_id,
grain_data_to_save=save_dict)
return grain_datafield