def gradient_xy(image):
# Compute the gradient of the image
grad = fast_gradient(image)
# Slice off the gradient for X and Y separately
grad_y = Image(grad.pixels[:image.n_channels])
grad_x = Image(grad.pixels[image.n_channels:])
return grad_x, grad_y
def gradient(self, image):
# Compute the gradient of the image
grad = fast_gradient(image)
# Create gradient image for X and Y
grad_y = Image(grad.pixels[: self.n_channels])
grad_x = Image(grad.pixels[self.n_channels :])
return grad_x, grad_y
def gradient(self, image):
# Compute the gradient of the image
grad = fast_gradient(image)
# Create gradient image for X and Y
grad_y = Image(grad.pixels[:self.n_channels])
grad_x = Image(grad.pixels[self.n_channels:])
return grad_x, grad_y
def gradient(self, image):
r"""
Function that computes the gradient of the image.
Parameters
----------
image : :map:`Image`
The input image.
Returns
-------
gradient : `ndarray`
The computed gradient.
"""
pixels = image.pixels
nabla = fast_gradient(pixels.reshape((-1, ) + self.patch_shape))
# remove 1st dimension gradient which corresponds to the gradient
# between parts
return nabla.reshape((2, ) + pixels.shape)
def gradient(self, image):
r"""
Function that computes the gradient of the image.
Parameters
----------
image : :map:`Image`
The input image.
Returns
-------
gradient : `ndarray`
The computed gradient.
"""
pixels = image.pixels
nabla = fast_gradient(pixels.reshape((-1,) + self.patch_shape))
# remove 1st dimension gradient which corresponds to the gradient
# between parts
return nabla.reshape((2,) + pixels.shape)
def gradient(self, image):
g = fast_gradient(image.pixels.reshape(
(-1,) + self.algorithm.appearance_model.mean().shape[-2:]))
return g.reshape((2,) + image.pixels.shape)
def gradient(self, image):
return fast_gradient(image).set_boundary_pixels().as_vector().reshape((2, image.n_channels, -1))
def gradient(self, img):
nabla = fast_gradient(img)
nabla.set_boundary_pixels()
return nabla.as_vector().reshape((2, img.n_channels, -1))
def gradient(self, image):
pixels = image.pixels
nabla = fast_gradient(pixels.reshape((-1,) + self.patch_shape))
# remove 1st dimension gradient which corresponds to the gradient
# between parts
return nabla.reshape((2,) + pixels.shape)