Name

recons3d_wbp - calculate 3-D reconstruction from a set of projections using weighted back-projection method (general or exact filter) or perform a simple backprojection

Usage

vol = recons3d_wbp(stack, list_proj, method [, symmetry, const])

Input

stack

name of the stack file or an in-core stack containing projection data, projections have to be squares, i.e., nx=ny.

list_proj

list of projections to be included in the reconstruction.

symmetry

point-group symmetry to be enforced, each projection will enter the reconstruction in all symmetry-related directions. The possibilities are

"cn"
"dn"
"i"
"c1" (no symmetry) is the default.

method

method used in WBP reconstruction, could either be general, exact, or empty string " ". The last indicates simple backprojection.

const

for general, it should correspond to the SNR of the reconstruction, the default is 1.0E4; for exact, it should be the diameter of the structure in pixels.

Output

vol: the 3-D reconstructed volume

Method

general: the weighting function is constructed as a sum of Gaussian function of distances between projection pixels in Fourier space, as described in: M. Radermacher, Weighted Back-Projection Methods. (p.102) In: Electron Tomography. Ed.: J. Frank, 1992.
exact: the weighting function is constructed as a sum of "overlaps" between Fourier transforms of projections, as described in: G. Harauz and M. van Heel, Exact filters for general geometry three dimensional reconstruction.Optik 73 (1986) 146-156. The smaller the diameter, the stronger the lowpass filtration effect on the reconstructed object.
" ": the projections are backprojected, no weighting applied

Author / Maintainer

Pawel A. Penczek

Keywords

category 1: INVERSION
category 2: FOURIER

Files

reconstruction.py

Maturity

alpha

Bugs

None. It is perfect.