def test_single_array_rgb(self):
"""Test that it can import just a single batch of images.
This is very useful for importing just testing data, where we do
not need segmentation maps. RGB version.
"""
x = utils.extract_data(
"tests/example_dataset/images_prepped_train/*.png", rgb=True)
x = utils.split_images(x, size=(60, 80), num_part=6)
expected_shape_x = (30, 60, 80, 3)
assert x.shape == expected_shape_x
def test_provide_label_path(self):
"""Test that it can import just a single batch of images.
This is very useful for importing just testing data, where we do
not need segmentation maps.
"""
x = utils.extract_data("tests/example_dataset/1.png", rgb=False)
expected_shape_x = (1, 512, 512, 1)
assert x.shape == expected_shape_x
def test_normalized(self):
"""Test that the images are actually normalized within
range 0-1. Very important!!!
"""
# This is the same image in grayscale
x, y = utils.extract_data("tests/example_dataset/1.png",
"tests/example_dataset/1.png",
rgb=False)
assert (x >= 0.0).all() and (x <= 1.0).all()
assert (y >= 0.0).all() and (y <= 1.0).all()
def test_grayscale(self):
"""Test that the extracted data have the correct color
channels. Very important!!!!
"""
# This is the same image in grayscale
x, y = utils.extract_data("tests/example_dataset/1.png",
"tests/example_dataset/1.png",
rgb=False)
expected_shape_x = (1, 512, 512, 1)
expected_shape_y = (1, 512, 512, 1)
assert x.shape == expected_shape_x
assert y.shape == expected_shape_y
def test_correct_shape(self):
"""Test that the extracted arrays have the correct
shape.
"""
x, y = utils.extract_data(
"tests/example_dataset/images_prepped_train/*.png",
"tests/example_dataset/annotations_prepped_train/*.png",
rgb=True,
)
# There are five images, RGB and grayscale for the segmentation maps
expected_shape_x = (5, 360, 480, 3)
expected_shape_y = (5, 360, 480, 1)
assert x.shape == expected_shape_x
assert y.shape == expected_shape_y
def test_correct_shape_grayscale(self):
"""Test that the extracted arrays have the correct
shape when they are RGB and grayscale.
"""
# First set is RGB, second one is grayscale
x, y = utils.extract_data(
"tests/example_dataset/images_prepped_train/*.png",
"tests/example_dataset/annotations_prepped_train/*.png",
rgb=True,
)
x, y = utils.split_images(x, y, size=(60, 80), num_part=6)
expected_shape_x = (30, 60, 80, 3)
expected_shape_y = (30, 60, 80, 1)
assert x.shape == expected_shape_x
assert y.shape == expected_shape_y
def test_correct_shape_rgb(self):
"""Test that the extracted arrays have the correct
shape when they are RGB.
"""
# Both sets of images are RGB
x, y = utils.extract_data(
"tests/example_dataset/images_prepped_train/*.png",
"tests/example_dataset/images_prepped_train/*.png",
rgb=True,
)
x, y = utils.split_images(x, y, size=(60, 80), num_part=6)
# There are five images and RGB, so there is a total of 30 patches
expected_shape_x = (30, 60, 80, 3)
expected_shape_y = (30, 60, 80, 3)
assert x.shape == expected_shape_x
assert y.shape == expected_shape_y
def test_batch_size(self):
"""Test that the batch size defined is the actual
value obtained.
"""
x, y = utils.extract_data(
"tests/example_dataset/images_prepped_train/*.png",
"tests/example_dataset/annotations_prepped_train/*.png",
rgb=True,
)
gen = utils.image_mask_augmentation(x, y, batch_size=2)
# There are five images and RGB, so there is a total of 30 patches
expected_shape_x = (2, 360, 480, 3)
expected_shape_y = (2, 360, 480, 1)
x_gen, y_gen = next(gen)
assert x_gen.shape == expected_shape_x
assert y_gen.shape == expected_shape_y