Filtering

This category contains 7 operator(s).

class AddPoissonNoise

Add Poisson Noise to Tilt Images

Add Poisson noise to tilt series. The number of counts (N) per tilt image can be specified below. The signal-to-noise ratio is square root of N.

transform(self, dataset, N=25)
Parameters

  • N (int) - Number of counts per tilt image

    • default: 25

    • minimum: 0

Implementation

  • Python script: AddPoissonNoise.py

  • JSON descriptor: AddPoissonNoise.json

class DeconvolutionDenoise

Deconvolution Denoise

Deconvolution-based denoising of volumetric data using a probe and a selected regularization method.

transform(self, dataset, probe=None, selected_scalars=None, method='APG_BM3D', fast_axis_scanning='dilation', slow_axis_scanning='average', max_iter=8, scale_x=1, scale_y=1, mu=1.0, axis=2, probe_kernel=11)
Parameters

  • selected_scalars (select_scalars)

  • method (enumeration) - The deconvolution method to use

    • default: 'APG_BM3D'

    • options:

      • 'APG_BM3D' - APG_BM3D

      • 'APG_TV' - APG_TV

      • 'ADMM_TV' - ADMM_TV

  • axis (enumeration) - The axis along which to process slices

    • default: 2

    • options:

      • 0 - X

      • 1 - Y

      • 2 - Z

  • probe (dataset) - The probe dataset used for deconvolution

  • fast_axis_scanning (enumeration) -

    • default: 'dilation'

    • options:

      • 'dilation' - dilation

      • 'average' - average

  • slow_axis_scanning (enumeration) -

    • default: 'average'

    • options:

      • 'dilation' - dilation

      • 'average' - average

  • probe_kernel (int) - The size of the probe kernel

    • default: 11

    • minimum: 1

  • scale_x (int) - X Scale factor

    • default: 1

    • minimum: 1

    • enabled if: method != 'ADMM_TV'

  • scale_y (int) - Y Scale factor

    • default: 1

    • minimum: 1

    • enabled if: method != 'ADMM_TV'

  • max_iter (int) - Maximum number of iterations

    • default: 8

    • minimum: 1

    • maximum: 100000

  • mu (double) - mu parameter

    • default: 0.01

    • precision: 5 decimal places

Implementation

  • Python script: DeconvolutionDenoise.py

  • JSON descriptor: DeconvolutionDenoise.json

class GaussianFilter

Gaussian Blur

Apply an isotropic Gaussian filter to 3D volume. The standard deviation (sigma) can be specified below:

transform(dataset, sigma=2.0)
Parameters

  • sigma (double)

    • default: 2.0

    • minimum: 0.0

Implementation

  • Python script: GaussianFilter.py

  • JSON descriptor: GaussianFilter.json

class GaussianFilterTiltSeries

Gaussian Filter

Apply a 2D isotropic Gaussian filter to each tilt image. The standard deviation (sigma) can be specified below:

transform(dataset, sigma=2.0)
Parameters

  • sigma (double)

    • default: 2.0

    • minimum: 0.0

Implementation

  • Python script: GaussianFilterTiltSeries.py

  • JSON descriptor: GaussianFilterTiltSeries.json

class MedianFilter

Median Filter

Apply an isotropic median filter. The window size can be specified below:

transform(dataset, size=2)
Parameters

  • size (int)

    • default: 2

    • minimum: 1

Implementation

  • Python script: MedianFilter.py

  • JSON descriptor: MedianFilter.json

class TV_Filter

TV_Filter

Apply a Total-Variation Minizimation-based Filter on tiff stacks, to remove structured artifacts (i.e. scratches, curtaining, or stripes) that span an angular range in Fourier Space. The range of intensities will be removed with a ‘Missing Wedge’ in Fourier Space and the lost information will be recovered with TV-Min

transform(self, dataset, Niter=100, a=0.1, wedgeSize=5, kmin=5, theta=0)
Parameters

  • wedgeSize (double)

    • default: 5.0

    • minimum: 0

    • maximum: 360

  • kmin (double)

    • default: 5.0

    • minimum: 0.0

    • maximum: 2000

  • theta (double)

    • default: 0.0

    • minimum: 0.0

  • Niter (int)

    • default: 50

    • minimum: 1

  • a (double)

    • default: 0.1

    • minimum: 0.0

    • maximum: 1.0

Implementation

  • Python script: TV_Filter.py

  • JSON descriptor: TV_Filter.json

class WienerFilter

Wiener Filter

Apply a Wiener filter to 3D volumetric dataset or stack of STEM projections. The standard deviations can be specified in each direction (x, y, z).

transform(dataset, SX=0.5, SY=0.5, SZ=0.5, noise=15.0)
Parameters

  • SX (double)

    • default: 0.5

    • minimum: 0.001

    • maximum: 1.0

  • SY (double)

    • default: 0.5

    • minimum: 0.001

    • maximum: 1.0

  • SZ (double)

    • default: 0.5

    • minimum: 0.001

    • maximum: 1.0

  • noise (double)

    • default: 15.0

    • minimum: 0.0

Implementation

  • Python script: WienerFilter.py

  • JSON descriptor: WienerFilter.json