def _image_projective_transform_grad(op, grad):
"""Computes the gradient for ImageProjectiveTransform."""
images = op.inputs[0]
transforms = op.inputs[1]
interpolation = op.get_attr("interpolation")
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(
transforms, name="transforms", dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# Invert transformations
transforms = flat_transforms_to_matrices(transforms=transforms)
inverse = linalg_ops.matrix_inverse(transforms)
transforms = matrices_to_flat_transforms(inverse)
output = gen_image_ops.image_projective_transform(
images=grad,
transforms=transforms,
output_shape=array_ops.shape(image_or_images)[1:3],
interpolation=interpolation)
return [output, None, None]
def transform(images, transforms, interpolation="NEAREST", name=None):
"""Applies the given transform(s) to the image(s).
Args:
images: A tensor of shape (num_images, num_rows, num_columns, num_channels)
(NHWC), (num_rows, num_columns, num_channels) (HWC), or
(num_rows, num_columns) (HW). The rank must be statically known (the
shape is not `TensorShape(None)`.
transforms: Projective transform matrix/matrices. A vector of length 8 or
tensor of size N x 8. If one row of transforms is
[a0, a1, a2, b0, b1, b2, c0, c1], then it maps the *output* point
`(x, y)` to a transformed *input* point
`(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
where `k = c0 x + c1 y + 1`. The transforms are *inverted* compared to
the transform mapping input points to output points. Note that gradients
are not backpropagated into transformation parameters.
interpolation: Interpolation mode. Supported values: "NEAREST", "BILINEAR".
Returns:
Image(s) with the same type and shape as `images`, with the given
transform(s) applied. Transformed coordinates outside of the input image
will be filled with zeros.
Raises:
TypeError: If `image` is an invalid type.
"""
with ops.name_scope(name, "transform"):
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(
transforms, name="transforms", dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
elif image_or_images.get_shape().ndims is None:
raise TypeError("image_or_images rank must be statically known")
elif len(image_or_images.get_shape()) == 2:
images = image_or_images[None, :, :, None]
elif len(image_or_images.get_shape()) == 3:
images = image_or_images[None, :, :, :]
elif len(image_or_images.get_shape()) == 4:
images = image_or_images
else:
raise TypeError("Images should have rank between 2 and 4.")
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif transform_or_transforms.get_shape().ndims is None:
raise TypeError(
"transform_or_transforms rank must be statically known")
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
output = gen_image_ops.image_projective_transform(
images, transforms, interpolation=interpolation.upper())
if len(image_or_images.get_shape()) == 2:
return output[0, :, :, 0]
elif len(image_or_images.get_shape()) == 3:
return output[0, :, :, :]
else:
return output
def _image_projective_transform_grad(op, grad):
"""Computes the gradient for ImageProjectiveTransform."""
images = op.inputs[0]
transforms = op.inputs[1]
interpolation = op.get_attr("interpolation")
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(transforms,
name="transforms",
dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# Invert transformations
transforms = flat_transforms_to_matrices(transforms=transforms)
inverse = linalg_ops.matrix_inverse(transforms)
transforms = matrices_to_flat_transforms(inverse)
output = gen_image_ops.image_projective_transform(
images=grad,
transforms=transforms,
output_shape=array_ops.shape(image_or_images)[1:3],
interpolation=interpolation)
return [output, None, None]
def test_projective_transform_v1(self):
"""The original ImageProjectiveTransform op should take 2 arguments."""
image = constant_op.constant([[[[1], [0]], [[0], [1]]]])
transform = constant_op.constant([[1., 0., 0., 0., 1., 0., 0., 0.]])
result = gen_image_ops.image_projective_transform(
image, transform, interpolation="NEAREST")
with self.cached_session():
self.assertAllEqual([[[[1], [0]], [[0], [1]]]], result.eval())
def test_projective_transform_v1(self):
"""The original ImageProjectiveTransform op should take 2 arguments."""
image = constant_op.constant([[[[1], [0]], [[0], [1]]]])
transform = constant_op.constant([[1., 0., 0., 0., 1., 0., 0., 0.]])
result = gen_image_ops.image_projective_transform(
image, transform, interpolation="NEAREST")
with self.cached_session():
self.assertAllEqual([[[[1], [0]], [[0], [1]]]], result.eval())
def transform(images, transforms):
"""Applies the given transform(s) to the image(s).
Args:
images: A tensor of shape (num_images, num_rows, num_columns, num_channels)
(NHWC), (num_rows, num_columns, num_channels) (HWC), or
(num_rows, num_columns) (HW).
transforms: Projective transform matrix/matrices. A vector of length 8 or
tensor of size N x 8. If one row of transforms is
[a0, a1, a2, b0, b1, b2, c0, c1], then it maps the *output* point
`(x, y)` to a transformed *input* point
`(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
where `k = c0 x + c1 y + 1`. The transforms are *inverted* compared to
the transform mapping input points to output points.
Returns:
Image(s) with the same type and shape as `images`, with the given
transform(s) applied. Transformed coordinates outside of the input image
will be filled with zeros.
Raises:
TypeError: If `image` is an invalid type.
"""
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(transforms,
name="transforms",
dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(image_or_images.get_shape()) == 2:
images = image_or_images[None, :, :, None]
elif len(image_or_images.get_shape()) == 3:
images = image_or_images[None, :, :, :]
elif len(image_or_images.get_shape()) == 4:
images = image_or_images
else:
raise TypeError("Images should have rank between 2 and 4.")
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# pylint: disable=protected-access
output = gen_image_ops.image_projective_transform(images, transforms)
if len(image_or_images.get_shape()) == 2:
return output[0, :, :, 0]
elif len(image_or_images.get_shape()) == 3:
return output[0, :, :, :]
else:
return output
def transform(images, transforms):
"""Applies the given transform(s) to the image(s).
Args:
images: A tensor of shape (num_images, num_rows, num_columns, num_channels)
(NHWC), (num_rows, num_columns, num_channels) (HWC), or
(num_rows, num_columns) (HW).
transforms: Projective transform matrix/matrices. A vector of length 8 or
tensor of size N x 8. If one row of transforms is
[a0, a1, a2, b0, b1, b2, c0, c1], then it maps the *output* point
`(x, y)` to a transformed *input* point
`(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
where `k = c0 x + c1 y + 1`. The transforms are *inverted* compared to
the transform mapping input points to output points.
Returns:
Image(s) with the same type and shape as `images`, with the given
transform(s) applied. Transformed coordinates outside of the input image
will be filled with zeros.
Raises:
TypeError: If `image` is an invalid type.
"""
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(
transforms, name="transforms", dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(image_or_images.get_shape()) == 2:
images = image_or_images[None, :, :, None]
elif len(image_or_images.get_shape()) == 3:
images = image_or_images[None, :, :, :]
elif len(image_or_images.get_shape()) == 4:
images = image_or_images
else:
raise TypeError("Images should have rank between 2 and 4.")
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# pylint: disable=protected-access
output = gen_image_ops.image_projective_transform(images, transforms)
if len(image_or_images.get_shape()) == 2:
return output[0, :, :, 0]
elif len(image_or_images.get_shape()) == 3:
return output[0, :, :, :]
else:
return output
def transform_4d(images, transforms, interpolation="NEAREST"):
""" A variant of tensorflow.contrib.image.python.ops.image_ops.rotate.transform
that only works on 4D tensors, without checking image rank.
when the shape of images is unknown, len(image_or_images.get_shape()) will raise error.
Applies the given transform(s) to the image(s).
Args:
images: A tensor of shape (num_images, num_rows, num_columns, num_channels)
(NHWC), (num_rows, num_columns, num_channels) (HWC), or
(num_rows, num_columns) (HW).
transforms: Projective transform matrix/matrices. A vector of length 8 or
tensor of size N x 8. If one row of transforms is
[a0, a1, a2, b0, b1, b2, c0, c1], then it maps the *output* point
`(x, y)` to a transformed *input* point
`(x', y') = ((a0 x + a1 y + a2) / k, (b0 x + b1 y + b2) / k)`,
where `k = c0 x + c1 y + 1`. The transforms are *inverted* compared to
the transform mapping input points to output points.
interpolation: Interpolation mode. Supported values: "NEAREST", "BILINEAR".
Returns:
Image(s) with the same type and shape as `images`, with the given
transform(s) applied. Transformed coordinates outside of the input image
will be filled with zeros.
Raises:
TypeError: If `image` is an invalid type.
"""
images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(transforms,
name="transforms",
dtype=dtypes.float32)
if images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % images.dtype)
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
output = gen_image_ops.image_projective_transform(
images, transforms, interpolation=interpolation.upper())
return output
def _image_projective_transform_grad(op, grad):
images = op.inputs[0]
transforms = op.inputs[1]
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(transforms,
name="transforms",
dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(image_or_images.get_shape()) == 2:
images = image_or_images[None, :, :, None]
elif len(image_or_images.get_shape()) == 3:
images = image_or_images[None, :, :, :]
elif len(image_or_images.get_shape()) == 4:
images = image_or_images
else:
raise TypeError("Images should have rank between 2 and 4")
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# Invert transformations
transforms = _flat_transforms_to_matrices(transforms=transforms)
inverse = linalg_ops.matrix_inverse(transforms)
transforms = _transform_matrices_to_flat(inverse)
output = gen_image_ops.image_projective_transform(grad, transforms)
if len(image_or_images.get_shape()) == 2:
return [output[0, :, :, 0], None]
elif len(image_or_images.get_shape()) == 3:
return [output[0, :, :, :], None]
else:
return [output, None]
def _image_projective_transform_grad(op, grad):
images = op.inputs[0]
transforms = op.inputs[1]
image_or_images = ops.convert_to_tensor(images, name="images")
transform_or_transforms = ops.convert_to_tensor(
transforms, name="transforms", dtype=dtypes.float32)
if image_or_images.dtype.base_dtype not in _IMAGE_DTYPES:
raise TypeError("Invalid dtype %s." % image_or_images.dtype)
if len(image_or_images.get_shape()) == 2:
images = image_or_images[None, :, :, None]
elif len(image_or_images.get_shape()) == 3:
images = image_or_images[None, :, :, :]
elif len(image_or_images.get_shape()) == 4:
images = image_or_images
else:
raise TypeError("Images should have rank between 2 and 4")
if len(transform_or_transforms.get_shape()) == 1:
transforms = transform_or_transforms[None]
elif len(transform_or_transforms.get_shape()) == 2:
transforms = transform_or_transforms
else:
raise TypeError("Transforms should have rank 1 or 2.")
# Invert transformations
transforms = _flat_transforms_to_matrices(transforms=transforms)
inverse = linalg_ops.matrix_inverse(transforms)
transforms = _transform_matrices_to_flat(inverse)
output = gen_image_ops.image_projective_transform(grad, transforms)
if len(image_or_images.get_shape()) == 2:
return [output[0, :, :, 0], None]
elif len(image_or_images.get_shape()) == 3:
return [output[0, :, :, :], None]
else:
return [output, None]
