cbi_toolbox.simu.primitives

The primitives module generates basic 3D objects

Conventions:

arrays follow the ZXY convention, with

  • Z : depth axis (axial, focus axis)

  • X : horizontal axis (lateral)

  • Y : vertical axis (lateral, rotation axis when relevant)

cbi_toolbox.simu.primitives.ball(size, radius=None, in_radius=0, antialias=2, dtype=<class 'numpy.float64'>)[source]

Generate hollow ball.

Parameters:

  • size (int) – side of the cube containing the volume

  • radius (float, optional) – radius of the ball, by default None (size / 2 - 1)

  • in_radius (float, optional) – radius of the inner (empty) sphere expressed as a fraction of the outer radius [0-1], by default 0 (plain ball)

  • antialias (int, optional) – antialiasing scale factor, by default 2

  • dtype (numpy.dtype or str, optional) – the datatype, by default numpy.float64

Returns:

The volume containing the ball.

Return type:

numpy.ndarray

Raises:

  • ValueError – if the size is odd

  • ValueError – if the antialias is negative

cbi_toolbox.simu.primitives.boccia(size, radius=None, n_stripes=3, deg_space=15, deg_width=7.5, rad_thick=0.12, antialias=2, dtype=<class 'numpy.float64'>)[source]

Create a boccia simulated sample: resolution stripes on a sphere

Parameters:

  • size (int) – side of the cube containing the volume (must be even)

  • radius (float, optional) – radius of the boccia, by default None (will be size / 2 - 1)

  • n_stripes (int, optional) – number of stripes to generate, by default 3

  • deg_space (int, optional) – spacing in degrees between the center of the stripes, by default 15

  • deg_width (float, optional) – width in degrees of the stripes, by default 7.5

  • rad_thick (float, optional) – thickness of the stripes, as a proportion of the radius, by default 0.12

  • antialias (int, optional) – antialiasing scale factor, by default 2

  • dtype (numpy.dtype or str, optional) – the datatype, by default numpy.float64

Returns:

The volume containing the boccia.

Return type:

numpy.ndarray

Raises:

  • ValueError – if the size is odd

  • ValueError – if the antialias is negative

  • ValueError – if the angle surpasses 90°

cbi_toolbox.simu.primitives.fiber(size, point, direction, radius, section='circle', antialias=2, dtype=<class 'numpy.float64'>)[source]
cbi_toolbox.simu.primitives.forward_ellipse(coordinates, center, radius, thickness=0.1, smoothing=0.5, solid=False)[source]

Compute a smoothed ellipse curve on the given coordinates.

Parameters:

  • coordinates (array) – the coordinates at which the ellipse is computed can be a meshgrid (3D array [2, w, h]) or a list of meshgrids (4D array [N, 2, w, h])

  • center (tuple (float, float)) – coordinates of the center of the ellipse

  • radius (tuple (float, float)) – radius of the ellipse

  • thickness (float, optional) – relative thickness of the ellipse curve, by default 0.1

  • smoothing (float, optional) – width of gaussian smoothing, by default 0.5

  • solid (bool, optional) – fill the ellipse, by default False

Returns:

the values of the ellipse at each point

Return type:

array

Raises:

  • ValueError – if the coordinates are in an invalid format

  • NotImplementedError – if non 2D ellipses are asked

cbi_toolbox.simu.primitives.phantom(size, scale=1, antialias=2)[source]

Generate 3D modified Shepp-Logann phantoms

Parameters:

  • size (int) – size of the output

  • scale (float) – scaling factor for the size of the phantoms

  • antialias (int, optional) – antialiasing factor, by default 2

Returns:

the phantoms

Return type:

array [ZXY]

Raises:

ValueError – if antialias is not a positive integer

cbi_toolbox.simu.primitives.quadrant_symmetry(quadrant)[source]

Generate a binary quadrant by rotating itself 90° on all 3 axes and summing.

Parameters:

quadrant (numpy.ndarray(bool)) – The original 3D quadrant.

Returns:

The symmetrized quadrant. Will have cubic shape equal to the largest dimension of the original quadrant.

Return type:

numpy.ndarray

cbi_toolbox.simu.primitives.quadrant_to_volume(quadrant, odd=(False, False, False))[source]

Generate a volume by mirroring a quadrant in all 8 corners.

Parameters:

  • quadrant (numpy.ndarray) – The quadrant corresponding to the end of the axes.

  • odd (tuple, optional) – If the target dimensions are odd, by default (False, False, False). If even, the dimension will be 2 * quandrant.shape. If odd, the dimension will be 2 * quadrant.shape - 1.

Returns:

The full volume.

Return type:

numpy.ndarray

cbi_toolbox.simu.primitives.torus_boccia(size, radius=None, n_stripes=3, deg_space=15, torus_radius=0.075, antialias=2, dtype=<class 'numpy.float64'>)[source]

Generate a boccia with torus stripes

Parameters:

  • size (int) – side of the cube containing the volume (must be even)

  • radius (float, optional) – radius of the boccia, by default None (size / 2 - 1)

  • n_stripes (int, optional) – number of torus stripes, by default 3

  • deg_space (int, optional) – spacing of the torus on the sphere in degrees, by default 15

  • torus_radius (float, optional) – radius of the torus, as a fraction of the sphere radius, by default 0.075

  • antialias (int, optional) – antialiasing scale factor, by default 2

  • dtype (numpy.dtype or str, optional) – the datatype, by default numpy.float64

Returns:

The volume containing the torus boccia.

Return type:

numpy.ndarray

Raises:

  • ValueError – if the size is odd

  • ValueError – if the antialias is negative

  • ValueError – if the angle surpasses 90°