kernelkit.fdk#
- kernelkit.fdk(projections: Any, projection_geometry: Any, volume_geometry: VolumeGeometry, filter: Any = 'ramlak', **bp_kwargs)#
Feldkamp-Davis-Kress algorithm for conebeam geometry.
- Parameters:
- projectionsarray-like
The projections to reconstruct from.
- projection_geometrylist[ProjectionGeometry]
The projection geometries.
- volume_geometryVolumeGeometry
The volume geometry.
- filterAny, optional
The filter to apply to the projections before backprojection. Default is ‘ramlak’. See kernelkit.processing.filter for more information.
- bp_kwargsdict, optional
Keyword arguments to pass to the bp function. See kernelkit.algo.bp for more information.
See also
kernelkit.algo.bp
Backprojection function.
Notes
The FDK algorithm is a special case of the bp function, with a VoxelDrivenConeBP kernel and a filter. See the article by Feldkamp et al. [1] for more information.
References
[1]Feldkamp, L. A., Davis, L. C., & Kress, J. W. (1984). Practical cone-beam algorithm. Journal of the Optical Society of America A, 1(6), 612. https://doi.org/10.1364/JOSAA.1.000612