def _crop_images_in_ram(self, filenames_and_boxes):
cropped_images = []
for filename, box in filenames_and_boxes:
cropped = load_and_crop_image(
filename, box, output_shape=self.input_shape)
bytes_io = io.BytesIO()
cropped.save(bytes_io, format='png')
cropped_images.append(bytes_io)
return cropped_images
def test_crop_box_is_used_with_resize_nearest(self):
# we crop to a 1 px image, and check that all image values
# are the same value
filename = get_loadable_filenames()[0]
box = (1, 1, 2, 2)
image = util.load_and_crop_image(filename, box)
correct_px_value = np.array(Image.open(filename))[box[0], box[1]]
self.assertTrue(np.all(np.array(image) == correct_px_value))
def read_image(self, filename, box):
image_pil = load_and_crop_image(filename,
box,
output_shape=self.input_shape)
image_np = np.array(image_pil).astype('float32')
image_np /= 255.0
image_np -= 0.5
image_np /= 0.25
image_torch = torch.from_numpy(image_np).repeat(3, 1,
1).to(self.device)
return image_torch
def _load_inputs_at_one_timepoint(self, f0_filename, box):
bottom, middle, top = augment_focus(f0_filename)
images = [
np.array(
load_and_crop_image(
name,
box,
output_shape=self.input_shape)
)
for name in [middle, top, bottom]]
images = np.array(images).astype('float32') / 255.0
images = (images - 0.5) / 0.25
return images
def test_output_image_is_correct_shape(self):
filename = get_loadable_filenames()[0]
box = (1, 1, 100, 100)
shape = (150, 140)
image = util.load_and_crop_image(filename, box, output_shape=shape)
self.assertEqual(image.size, shape)
def test_returns_pil_image(self):
filename = get_loadable_filenames()[0]
box = (1, 1, 2, 2)
image = util.load_and_crop_image(filename, box)
self.assertIsInstance(image, Image.Image)