Reconstruction

This category contains 4 operator(s).

class Recon_DFT

Direct Fourier Reconstruction

Reconstruct a tilt series using Direct Fourier Method (DFM). The tilt axis must be parallel to the x-direction and centered in the y-direction. The size of reconstruction will be (Nx,Ny,Ny). Reconstrucing a 512x512x512 tomogram typically takes 30-40 seconds.

Note

This operator is not compatible with external Python execution.

transform(self, dataset)

Child Datasets

• reconstruction (reconstruction) - Reconstruction

Implementation

• Python script: Recon_DFT.py

• JSON descriptor: Recon_DFT.json

class Recon_DFT_constraint

Reconstruct (Constraint based Direct Fourier)

Reconstruct a tilt series using constraint-based Direct Fourier method. The tilt axis must be parallel to the x-direction and centered in the y-direction. The size of reconstruction will be (Nx,Ny,Ny). Reconstructing a 512x512x512 tomogram typically takes xxxx mins.

Note

This operator is not compatible with external Python execution.

transform(self, dataset, Niter=None, Niter_update_support=None, supportSigma=None, supportThreshold=None, seed=0)

Parameters

• Niter (int)

  • default: 100

  • minimum: 1

  • maximum: 10000

• Niter_update_support (int)

  • default: 50

  • minimum: 0

  • maximum: 10000

  • step: 10

• supportSigma (double)

  • default: 0.1

  • minimum: 0

  • maximum: 1

  • step: 0.1

• supportThreshold (double)

  • default: 10.0

  • minimum: 0

  • maximum: 100

  • step: 5

• seed (int) - Seed for the random number generator used to create the initial guess. The seed is kept here so that when the pipeline is re-ran, the results are reproducible

  • default: 0

  • minimum: 0

  • maximum: 1000000000

Child Datasets

• reconstruction (reconstruction) - Reconstruction

Implementation

• Python script: Recon_DFT_constraint.py

• JSON descriptor: Recon_DFT_constraint.json

class Recon_SIRT

SIRT Reconstruction

Reconstruct a tilt series using Simultaneous Iterative Reconstruction Techniques Technique (SIRT) with a Positivity Constraint. The tilt series data should be aligned prior to reconstruction and the tilt axis must be parallel to the x-direction. The size of reconstruction will be (Nx,Ny,Ny). The number of iterations can be specified below. Reconstrucing a 256x256x256 tomogram typically with Landweber’s Method takes about 2 mins for 10 iterations.

transform(self, dataset, Niter=10, stepSize=0.0001, updateMethodIndex=0, Nupdates=0)

Parameters

• Niter (int)

  • default: 10

  • minimum: 1

• stepSize (double)

  • default: 0.0001

  • minimum: 0

  • precision: 5 decimal places

  • step: 0.0001

• updateMethodIndex (enumeration)

  • default: 0

  • options:

    • 0 - Landweber

    • 1 - Cimmino

    • 2 - Component average

• Nupdates (int)

  • default: 0

  • minimum: 0

  • maximum: 100

Child Datasets

• reconstruction (reconstruction) - Reconstruction

Implementation

• Python script: Recon_SIRT.py

• JSON descriptor: Recon_SIRT.json

class Recon_tomopy

TomoPy Reconstruction

Run TomoPy reconstruction on a dataset

transform(dataset, algorithm='gridrec', num_iter=5)

Parameters

• algorithm (enumeration) - Reconstruction algorithm to use

  • default: 'gridrec'

  • options:

    • 'gridrec' - Gridrec

    • 'fbp' - FBP

    • 'mlem' - MLEM

    • 'ospml_hybrid' - OSPML Hybrid

• num_iter (int) - Number of iterations (iterative methods only)

  • default: 5

  • minimum: 1

  • maximum: 1000

  • visible if: algorithm == 'mlem' or algorithm == 'ospml_hybrid'

Child Datasets

• reconstruction (reconstruction) - Reconstruction

Implementation

• Python script: Recon_tomopy.py

• JSON descriptor: Recon_tomopy.json