def test_copysegmentation_dvid_to_zarr(setup_dvid_to_zarr):
template_dir, config, volume, dvid_address, repo_uuid, output_file = setup_dvid_to_zarr
# Modify the config from above to compute pyramid scales,
# and choose a bounding box that is aligned with the bricks even at scale 2
# (just for easier testing).
box_zyx = [[0, 0, 0], [256, 256, 256]]
config["input"]["geometry"]["bounding-box"] = box_zyx
config["copysegmentation"]["pyramid-depth"] = 2
yaml = YAML()
yaml.default_flow_style = False
with open(f"{template_dir}/workflow.yaml", 'w') as f:
yaml.dump(config, f)
execution_dir, _workflow = launch_flow(template_dir, 1)
box_zyx = np.array(box_zyx)
scale_0_vol = volume[box_to_slicing(*box_zyx)]
scale_1_vol = downsample_labels(scale_0_vol, 2, True)
scale_2_vol = downsample_labels(scale_1_vol, 2, True)
store = zarr.NestedDirectoryStore(f"{execution_dir}/{output_file}")
f = zarr.open(store, 'r')
output_0_vol = f['s0'][box_to_slicing(*(box_zyx // 1))]
output_1_vol = f['s1'][box_to_slicing(*(box_zyx // 2))]
output_2_vol = f['s2'][box_to_slicing(*(box_zyx // 4))]
assert (output_0_vol == scale_0_vol).all(), \
"Scale 0: Written vol does not match expected"
assert (output_1_vol == scale_1_vol).all(), \
"Scale 1: Written vol does not match expected"
assert (output_2_vol == scale_2_vol).all(), \
"Scale 2: Written vol does not match expected"
def test_copygrayscale_from_hdf5_to_dvid_multiscale(setup_hdf5_grayscale_input,
disable_auto_retry):
_template_dir, config, _volume, dvid_address, repo_uuid, output_grayscale_name = setup_hdf5_grayscale_input
# Modify the config from above to compute pyramid scales,
# and choose a bounding box that is aligned with the bricks even at scale 2
# (just for easier testing).
config["input"]["geometry"]["bounding-box"] = [[0, 0, 0], [128, 256, 128]]
config["copygrayscale"]["min-pyramid-scale"] = 1
config["copygrayscale"]["max-pyramid-scale"] = 2
config["copygrayscale"][
"downsample-method"] = "label" # This test is easier to write if we use this downsampling method
box_zyx, scale_0_vol = _run_to_dvid(setup_hdf5_grayscale_input,
check_scale_0=False)
scale_1_vol = downsample_labels(scale_0_vol, 2, True)
scale_2_vol = downsample_labels(scale_1_vol, 2, True)
# Check the other scales -- be careful to extract exactly one brick.
output_vol = fetch_raw(dvid_address, repo_uuid,
output_grayscale_name + '_1', box_zyx // 2)
assert (output_vol == scale_1_vol).all(), \
"Scale 1: Written vol does not match expected"
# Check the other scales -- be careful to extract exactly one brick.
output_vol = fetch_raw(dvid_address, repo_uuid,
output_grayscale_name + '_2', box_zyx // 4)
assert (output_vol == scale_2_vol).all(), \
"Scale 2: Written vol does not match expected"
def downsample(volume, factor, method):
assert method in DOWNSAMPLE_METHODS
assert (np.array(volume.shape) % factor == 0).all(), \
"Volume dimensions must be a multiple of the downsample factor."
if method == 'subsample':
sl = slice(None, None, factor)
return volume[(sl, ) * volume.ndim].copy('C')
if method == 'block-mean':
return block_downsample(volume, factor)
if method in ('zoom', 'grayscale'): # synonyms
# vigra is 2.7x faster than scipy, but it complains for small images:
#
# Precondition violation!
# resizeImage(): Each output axis must have length > 1.)
#
# Furthermore, it seems to be somewhat unstable even for small
# images that satisfy the above condition.
# Therefore, use vigra for non-tiny images,
# and use scipy for the tiny stuff.
newshape = np.array(volume.shape) // factor
if (newshape >= 10).all():
return vigra_sampling_resize(volume, newshape)
else:
return scipy.ndimage.zoom(volume, 1 / factor)
if method == 'mode':
return downsample_labels(volume, factor, False)
if method in ('labels', 'label'): # synonyms
return downsample_labels(volume, factor, True)
if method == 'labels-numba':
reduced_output, _reduced_box = downsample_labels_3d_suppress_zero(
volume, factor)
return reduced_output
raise AssertionError("Shouldn't get here.")
def downsample_brick(brick):
assert (brick.physical_box % factor == 0).all()
assert (brick.logical_box % factor == 0).all()
if method == 'grayscale':
downsampled_volume = scipy.ndimage.interpolation.zoom(brick.volume, 1/factor, mode='reflect')
elif method == 'label':
# Old: Python downsampling
# downsample_3Dlabels(brick.volume)
# Newer: Numba downsampling
#downsampled_volume, _ = downsample_labels_3d_suppress_zero(brick.volume, (2,2,2), brick.physical_box)
# Even Newer: C++ downsampling (note: only works on aligned data.)
# For consistency with DVID's on-demand downsampling, we suppress 0 pixels.
downsampled_volume = downsample_labels(brick.volume, factor, suppress_zero=True)
else:
assert False
downsampled_logical_box = brick.logical_box // factor
downsampled_physical_box = brick.physical_box // factor
return Brick(downsampled_logical_box, downsampled_physical_box, downsampled_volume)
def test_copysegmentation_from_hdf5_to_dvid_multiscale(
setup_hdf5_segmentation_input, disable_auto_retry):
template_dir, config, volume, dvid_address, repo_uuid, _ = setup_hdf5_segmentation_input
# Modify the config from above to compute pyramid scales,
# and choose a bounding box that is aligned with the bricks even at scale 2
# (just for easier testing).
box_zyx = [[0, 0, 0], [256, 256, 256]]
config["input"]["geometry"]["bounding-box"] = box_zyx
config["copysegmentation"]["pyramid-depth"] = 2
# Change the segmentation name so it doesn't conflict with earlier tests
output_segmentation_name = 'segmentation-output-from-hdf5-multiscale'
config["output"]["dvid"]["segmentation-name"] = output_segmentation_name
yaml = YAML()
yaml.default_flow_style = False
with open(f"{template_dir}/workflow.yaml", 'w') as f:
yaml.dump(config, f)
_execution_dir, _workflow = launch_flow(template_dir, 1)
box_zyx = np.array(box_zyx)
scale_0_vol = volume[box_to_slicing(*box_zyx)]
scale_1_vol = downsample_labels(scale_0_vol, 2, True)
scale_2_vol = downsample_labels(scale_1_vol, 2, True)
output_0_vol = fetch_labelmap_voxels(dvid_address,
repo_uuid,
output_segmentation_name,
box_zyx // 1,
scale=0)
output_1_vol = fetch_labelmap_voxels(dvid_address,
repo_uuid,
output_segmentation_name,
box_zyx // 2,
scale=1)
output_2_vol = fetch_labelmap_voxels(dvid_address,
repo_uuid,
output_segmentation_name,
box_zyx // 4,
scale=2)
# np.save('/tmp/expected-0.npy', scale_0_vol)
# np.save('/tmp/expected-1.npy', scale_1_vol)
# np.save('/tmp/expected-2.npy', scale_2_vol)
#
# np.save('/tmp/output-0.npy', output_0_vol)
# np.save('/tmp/output-1.npy', output_1_vol)
# np.save('/tmp/output-2.npy', output_2_vol)
#
# np.save('/tmp/diff-0.npy', (output_0_vol != scale_0_vol))
# np.save('/tmp/diff-1.npy', (output_1_vol != scale_1_vol))
# np.save('/tmp/diff-2.npy', (output_2_vol != scale_2_vol))
assert (output_0_vol == scale_0_vol).all(), \
"Scale 0: Written vol does not match expected"
assert (output_1_vol == scale_1_vol).all(), \
"Scale 1: Written vol does not match expected"
assert (output_2_vol == scale_2_vol).all(), \
"Scale 2: Written vol does not match expected"