colourlab.image_core module

image_core: Colour image core operations, part of the colourlab package

For image processing operations that are applied directly to ndarrays. To be used by the colour-space-specific methods in colour.image.Image.

Copyright (C) 2017-2021 Ivar Farup

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see <http://www.gnu.org/licenses/>.

colourlab.image_core.correlate(im, filter)[source]

Correlation filter for multi-channel image with symmetric boundaries.

im : ndarray
M x N x C image

filter : ndarray

colourlab.image_core.diffusion_tensor_from_structure(s_tuple, dpsi_dlambda1=None, dpsi_dlambda2=None)[source]

Compute the diffusion tensor coefficients from the structure tensor parameters

Parameters:
  • s_tuple (tuple) – The resulting tuple from a call to image.structure_tensor
  • dpsi_dlambda1 (func) – The diffusion supression function for the first eigenvalue of the structure tensor. If None, use Perona and Malik’s inverse square with default value from image_core.
  • dpsi_dlambda2 (func) – Same for the second eigenvalue. If None use dpsi_dlambda1
Returns:

  • d11 (ndarray) – The d11 component of the structure tensor of the image data.
  • d12 (ndarray) – The d12 component of the structure tensor of the image data.
  • d22 (ndarray) – The d22 component of the structure tensor of the image data.
colourlab.image_core.divergence(imi, imj, diff=(array([[ 0. , -0.5, 0. ], [ 0. , 0. , 0. ], [ 0. , 0.5, 0. ]]), array([[ 0. , 0. , 0. ], [-0.5, 0. , 0.5], [ 0. , 0. , 0. ]])))[source]

Compute the divergence of the image components with the given filters.

Parameters:
  • imi (ndarray) – M x N x C image
  • imj (ndarray) – M x N x C image
  • diff (tuple) – Tuple with the two gradient filters
Returns:

ndarray
Return type:

div(imi, imj)
colourlab.image_core.dpdl_perona_exp(lambd, kappa=0.01)[source]

Perona and Malik’s exponential diffusion supression

Name suggest dpsi/dlambda

Parameters:
  • lambd (ndimage) – Eigenvalue of the structure tensor (single channelimage)
  • kappa (float) – Paramter
Returns:

ndimage
Return type:

dpsi / dlambda
colourlab.image_core.dpdl_perona_invsq(lambd, kappa=0.01)[source]

Perona and Malik’s inverse square diffusion supression

Name suggest dpsi/dlambda = 1 / (1 + lambd / kappa^2)

Parameters:
  • lambd (ndimage) – Eigenvalue of the structure tensor (single channelimage)
  • kappa (float) – Paramter
  • = 1 / (1 + lambd / kappa**2) (Returns) –
  • ------- –
  • ndimage (dpsi / dlambda) –
colourlab.image_core.dpdl_tv(lambd, epsilon=0.0001)[source]

Total variation diffusion supression

Name suggest dpsi/dlambda = 1 / sqrt(lambd + epsilon)

Parameters:
  • lambd (ndimage) – Eigenvalue of the structure tensor (single channelimage)
  • epsilon (float) – Regularisation paramter
Returns:

ndimage
Return type:

dpsi / dlambda
colourlab.image_core.gradient(im, diff=(array([[ 0. , -0.5, 0. ], [ 0. , 0. , 0. ], [ 0. , 0.5, 0. ]]), array([[ 0. , 0. , 0. ], [-0.5, 0. , 0.5], [ 0. , 0. , 0. ]])))[source]

Compute the gradient of the image with the given filters.

Parameters:
  • im (ndarray) – M x N x C image
  • diff (tuple) – Tuple with the two gradient filters
Returns:

  • ndarray (d im / di)
  • ndarray (d im / dj)
colourlab.image_core.jit(func)[source]
colourlab.image_core.stress(im, ns=3, nit=5, R=0)[source]

Compute the stress image and range for a multi-channel image.

Parameters:
  • im (ndarray) – Multi-channel image
  • ns (int) – Number of sample points
  • nit (int) – Number of iterations
  • R (int) – Maximum radius. If R=0, the diagonal of the image is used.
Returns:

  • stress_im (ndarray) – The result of stress
  • range_im (ndarray) – The range image (see paper)
colourlab.image_core.stress_channel(*args, **kwargs)[source]