CircularDetector¶

class odl.tomo.geometry.detector.CircularDetector(partition, axis, radius, check_bounds=True)[source]¶

Bases: Detector

A 1D detector on a circle section in 2D space.

The circular section that corresponds to the angular partition is rotated to be aligned with a given axis and shifted to cross the origin. Note, the partition angle increases in the clockwise direction, by analogy to flat detectors.

Attributes:

axis: Fixed axis along which this detector is aligned.
check_bounds: If True, methods computing vectors check input arguments.
grid: Sampling grid of the parameters.
ndim: Number of dimensions of the parameters (= surface dimension).
params: Surface parameter set of this detector.
partition: Partition of the detector parameter set into subsets.
radius: Curvature radius of the detector.
rotation_matrix: Rotation matrix that is used to align the detector with a given axis.
shape: Number of subsets (pixels) of the detector per axis.
size: Total number of pixels.
space_ndim: Number of dimensions of the embedding space.
translation: A vector used to shift the detector towards the origin.

Methods

`surface`(param)	Return the detector surface point corresponding to `param`.
`surface_deriv`(param)	Return the surface derivative at `param`.
`surface_measure`(param)	Return the arc length measure at `param`.
`surface_normal`(param)	Unit vector perpendicular to the detector surface at `param`.

__init__(partition, axis, radius, check_bounds=True)[source]¶

Initialize a new instance.

Parameters:

partition1-dim. RectPartition: Partition of the parameter interval, corresponding to the angular sections along the line.
axisarray-like, shape (2,): Fixed axis along which this detector is aligned.
radiusnonnegative float: Radius of the circle.
check_boundsbool, optional: If True, methods computing vectors check input arguments. Checks are vectorized and add only a small overhead.

Examples

Initialize a detector with circle radius 2 and extending to 90 degrees on both sides of the origin (a half circle).

>>> part = odl.uniform_partition(-np.pi / 2, np.pi / 2, 10)
>>> det = CircularDetector(part, axis=[1, 0], radius=2)
>>> det.axis
array([ 1.,  0.])
>>> det.radius
2.0
>>> np.allclose(det.surface_normal(0), [0, -1])
True