def get(self, image, sample, objective, **kwargs):
# Grab properties from the objective to pass to the sample
new_kwargs = objective.properties.current_value_dict()
new_kwargs.update(kwargs)
kwargs = new_kwargs
list_of_scatterers = sample.resolve(**kwargs)
if not isinstance(list_of_scatterers, list):
list_of_scatterers = [list_of_scatterers]
sample_volume, limits = _create_volume(list_of_scatterers, **kwargs)
sample_volume = Image(sample_volume)
for scatterer in list_of_scatterers:
sample_volume.merge_properties_from(scatterer)
imaged_sample = objective.resolve(sample_volume,
limits=limits,
**kwargs)
# Merge with input
if not image:
return imaged_sample
if not isinstance(image, list):
image = [image]
for i in range(len(image)):
image[i].merge_properties_from(imaged_sample)
return image
def get(self, images, axis, features, **kwargs):
if features is not None:
images = [feature.resolve() for feature in features]
result = Image(np.mean(images, axis=axis))
for image in images:
result.merge_properties_from(image)
return result
def get(self, image, sample, objective, **kwargs):
# Grab properties from the objective to pass to the sample
new_kwargs = objective.properties.current_value_dict(**kwargs)
new_kwargs.update(kwargs)
kwargs = new_kwargs
list_of_scatterers = sample.resolve(**kwargs)
if not isinstance(list_of_scatterers, list):
list_of_scatterers = [list_of_scatterers]
volume_samples = [
scatterer
for scatterer in list_of_scatterers
if not scatterer.get_property("is_field", default=False)
]
field_samples = [
scatterer
for scatterer in list_of_scatterers
if scatterer.get_property("is_field", default=False)
]
sample_volume, limits = _create_volume(volume_samples, **kwargs)
sample_volume = Image(sample_volume)
for scatterer in volume_samples + field_samples:
sample_volume.merge_properties_from(scatterer)
imaged_sample = objective.resolve(
sample_volume, limits=limits, fields=field_samples, **kwargs
)
upscale = kwargs["upscale"]
shape = imaged_sample.shape
if upscale > 1:
mean_imaged_sample = np.reshape(
imaged_sample,
(shape[0] // upscale, upscale, shape[1] // upscale, upscale, shape[2]),
).mean(axis=(3, 1))
imaged_sample = Image(mean_imaged_sample).merge_properties_from(
imaged_sample
)
# Merge with input
if not image:
return imaged_sample
if not isinstance(image, list):
image = [image]
for i in range(len(image)):
image[i].merge_properties_from(imaged_sample)
return image
def get(self, image, scale, translate, rotate, shear, **kwargs):
assert (
image.ndim == 2 or image.ndim == 3
), "Affine only supports 2-dimensional or 3-dimension inputs."
dx, dy = translate
fx, fy = scale
cr = np.cos(rotate)
sr = np.sin(rotate)
k = np.tan(shear)
scale_map = np.array([[1 / fx, 0], [0, 1 / fy]])
rotation_map = np.array([[cr, sr], [-sr, cr]])
shear_map = np.array([[1, 0], [-k, 1]])
mapping = scale_map @ rotation_map @ shear_map
shape = image.shape
center = np.array(shape[:2]) / 2
d = center - np.dot(mapping, center) - np.array([dy, dx])
# Clean up kwargs
kwargs.pop("input", False)
kwargs.pop("matrix", False)
kwargs.pop("offset", False)
kwargs.pop("output", False)
# Call affine_transform
if image.ndim == 2:
new_image = utils.safe_call(
ndimage.affine_transform,
input=image,
matrix=mapping,
offset=d,
**kwargs
)
new_image = Image(new_image)
new_image.merge_properties_from(image)
image = new_image
elif image.ndim == 3:
for z in range(shape[-1]):
image[:, :, z] = utils.safe_call(
ndimage.affine_transform,
input=image[:, :, z],
matrix=mapping,
offset=d,
**kwargs
)
# Map positions
inverse_mapping = np.linalg.inv(mapping)
for prop in image.properties:
if "position" in prop:
position = np.array(prop["position"])
prop["position"] = np.array(
(
*(
(
inverse_mapping
@ (position[:2] - center + np.array([dy, dx]))
+ center
)
),
*position[3:],
)
)
return image
def _process_and_get(self, images, **kwargs):
if isinstance(images, list) and len(images) != 1:
list_of_labels = super()._process_and_get(images, **kwargs)
else:
if isinstance(images, list):
images = images[0]
list_of_labels = []
for prop in images.properties:
if "position" in prop:
inp = Image(np.array(images))
inp.append(prop)
out = Image(self.get(inp, **kwargs))
out.merge_properties_from(inp)
list_of_labels.append(out)
output_region = kwargs["output_region"]
output = np.zeros(
(output_region[2], output_region[3], kwargs["number_of_masks"])
)
for label in list_of_labels:
positions = _get_position(label)
for position in positions:
p0 = np.round(position - output_region[0:2])
if np.any(p0 > output.shape[0:2]) or np.any(p0 + label.shape[0:2] < 0):
continue
crop_x = int(-np.min([p0[0], 0]))
crop_y = int(-np.min([p0[1], 0]))
crop_x_end = int(
label.shape[0]
- np.max([p0[0] + label.shape[0] - output.shape[0], 0])
)
crop_y_end = int(
label.shape[1]
- np.max([p0[1] + label.shape[1] - output.shape[1], 0])
)
labelarg = label[crop_x:crop_x_end, crop_y:crop_y_end, :]
p0[0] = np.max([p0[0], 0])
p0[1] = np.max([p0[1], 0])
p0 = p0.astype(np.int)
output_slice = output[
p0[0] : p0[0] + labelarg.shape[0], p0[1] : p0[1] + labelarg.shape[1]
]
for label_index in range(kwargs["number_of_masks"]):
if isinstance(kwargs["merge_method"], list):
merge = kwargs["merge_method"][label_index]
else:
merge = kwargs["merge_method"]
if merge == "add":
output[
p0[0] : p0[0] + labelarg.shape[0],
p0[1] : p0[1] + labelarg.shape[1],
label_index,
] += labelarg[..., label_index]
elif merge == "overwrite":
output_slice[
labelarg[..., label_index] != 0, label_index
] = labelarg[labelarg[..., label_index] != 0, label_index]
output[
p0[0] : p0[0] + labelarg.shape[0],
p0[1] : p0[1] + labelarg.shape[1],
label_index,
] = output_slice[..., label_index]
elif merge == "or":
output[
p0[0] : p0[0] + labelarg.shape[0],
p0[1] : p0[1] + labelarg.shape[1],
label_index,
] = (output_slice[..., label_index] != 0) | (
labelarg[..., label_index] != 0
)
elif merge == "mul":
output[
p0[0] : p0[0] + labelarg.shape[0],
p0[1] : p0[1] + labelarg.shape[1],
label_index,
] *= labelarg[..., label_index]
else:
# No match, assume function
output[
p0[0] : p0[0] + labelarg.shape[0],
p0[1] : p0[1] + labelarg.shape[1],
label_index,
] = merge(
output_slice[..., label_index], labelarg[..., label_index]
)
output = Image(output)
for label in list_of_labels:
output.merge_properties_from(label)
return output