ImageDataFilters.median_smooth

ImageDataFilters.median_smooth(

*,

kernel_size: VectorLikeInt = (3, 3, 3),

scalars: str | None = None,

preference: PointLiteral | CellLiteral = 'point',

progress_bar: bool = False,

) → ImageData

Smooth data using a median filter.

The Median filter that replaces each pixel with the median value from a rectangular neighborhood around that pixel. Neighborhoods can be no more than 3 dimensional. Setting one axis of the neighborhood kernelSize to 1 changes the filter into a 2D median.

See vtkImageMedian3D for more details.

Parameters:

kernel_sizeVectorLikeInt, default: (3, 3, 3)

Size of the kernel in each dimension (units of voxels), for example (x_size, y_size, z_size). Default is a 3D median filter. If you want to do a 2D median filter, set the size to 1 in the dimension you don’t want to filter over.

scalarsstr, optional

Name of scalars to process. Defaults to currently active scalars.

preferencestr, default: “point”

When scalars is specified, this is the preferred array type to search for in the dataset. Must be either 'point' or 'cell'.

progress_barbool, default: False

Display a progress bar to indicate progress.

Returns:

pyvista.ImageData

Uniform grid with smoothed scalars.

Warning

Applying this filter to cell data will send the output to a new point array with the same name, overwriting any existing point data array with the same name.

Examples

Download Python source code | Download Jupyter notebook

First, create sample data to smooth. Here, we use pyvista.perlin_noise() to create meaningful data.

>>> import numpy as np
>>> import pyvista as pv
>>> noise = pv.perlin_noise(0.1, (2, 5, 8), (0, 0, 0))
>>> grid = pv.sample_function(
...     noise, bounds=[0, 1, 0, 1, 0, 1], dim=(20, 20, 20)
... )
>>> grid.plot(show_scalar_bar=False)

Static Scene

Interactive Scene

Next, smooth the sample data.

>>> smoothed = grid.median_smooth(kernel_size=(10, 10, 10))
>>> smoothed.plot(show_scalar_bar=False)

Static Scene

Interactive Scene