def augment_resize(sample_data, sample_seg, target_size, order=3, order_seg=1, cval_seg=0):
"""
Reshapes data (and seg) to target_size
:param sample_data: np.ndarray or list/tuple of np.ndarrays, must be (c, x, y(, z))) (if list/tuple then each entry
must be of this shape!)
:param target_size: int or list/tuple of int
:param order: interpolation order for data (see skimage.transform.resize)
:param order_seg: interpolation order for seg (see skimage.transform.resize)
:param cval_seg: cval for segmentation (see skimage.transform.resize)
:param sample_seg: can be None, if not None then it will also be resampled to target_size. Can also be list/tuple of
np.ndarray (just like data). Must also be (c, x, y(, z))
:return:
"""
dimensionality = len(sample_data.shape) - 1
if not isinstance(target_size, (list, tuple)):
target_size_here = [target_size] * dimensionality
else:
assert len(target_size) == dimensionality, "If you give a tuple/list as target size, make sure it has " \
"the same dimensionality as data!"
target_size_here = list(target_size)
sample_data = resize_multichannel_image(sample_data, target_size_here, order)
if sample_seg is not None:
target_seg = np.ones([sample_seg.shape[0]] + target_size_here)
for c in range(sample_seg.shape[0]):
target_seg[c] = resize_segmentation(sample_seg[c], target_size_here, order_seg, cval_seg)
else:
target_seg = None
return sample_data, target_seg
def augment_zoom(sample_data, sample_seg, zoom_factors, order=3, order_seg=1, cval_seg=0):
"""
zooms data (and seg) by factor zoom_factors
:param sample_data: np.ndarray or list/tuple of np.ndarrays, must be (c, x, y(, z))) (if list/tuple then each entry
must be of this shape!)
:param zoom_factors: int or list/tuple of int (multiplication factor for the input size)
:param order: interpolation order for data (see skimage.transform.resize)
:param order_seg: interpolation order for seg (see skimage.transform.resize)
:param cval_seg: cval for segmentation (see skimage.transform.resize)
:param sample_seg: can be None, if not None then it will also be zoomed by zoom_factors. Can also be list/tuple of
np.ndarray (just like data). Must also be (c, x, y(, z))
:return:
"""
dimensionality = len(sample_data.shape) - 1
shape = np.array(sample_data.shape[1:])
if not isinstance(zoom_factors, (list, tuple)):
zoom_factors_here = np.array([zoom_factors] * dimensionality)
else:
assert len(zoom_factors) == dimensionality, "If you give a tuple/list as target size, make sure it has " \
"the same dimensionality as data!"
zoom_factors_here = np.array(zoom_factors)
target_shape_here = list(np.round(shape * zoom_factors_here).astype(int))
sample_data = resize_multichannel_image(sample_data, target_shape_here, order)
if sample_seg is not None:
target_seg = np.ones([sample_seg.shape[0]] + target_shape_here)
for c in range(sample_seg.shape[0]):
target_seg[c] = resize_segmentation(sample_seg[c], target_shape_here, order_seg, cval_seg)
else:
target_seg = None
return sample_data, target_seg
def resize_data_and_seg(data,
size,
seg=None,
order_data=3,
order_seg=1,
cval_seg=0):
"""
Args:
data (np.ndarray): shape (b, c, h, w (, d))
seg (np.ndarray): shape (b, c, h, w (, d)). Defaults to None.
size (list/tuple of int): size to resize to
does not include the batch size or number of channels
order_data (int): interpolation order for data
(see skimage.transform.resize)
order_seg (int): interpolation order for seg
(see skimage.transform.resize)
"""
target_data = np.ones(list(data.shape[:2]) + size)
if seg is not None:
target_seg = np.ones(list(seg.shape[:2]) + size)
else:
target_seg = None
for b in range(len(data)):
target_data[b] = resize_multichannel_image(data[b], size, order_data)
if seg is not None:
for c in range(seg.shape[1]):
target_seg[b, c] = resize_segmentation(seg[b, c], size,
order_seg, cval_seg)
return target_data, target_seg
