Results Filter Dialog¶

There are basically two different types of filtering that can be done, i.e. Spatial - Gauss Weighting and Time - Moving Average. These will be explained below. The other types available in the dialog box must be seen as experimental.

Spatial - Gauss Weighting

The Gauss filter is a spatial filter, i.e. it acts on the nodal displacements for a particular time step (frame id) and smooths the displacements in the image plane.

To define a filter, you must first specify the source. By default it is the Raw DIC Data, but it can also be any previosly filtered data.
The filter type is now set as Spatial - Gaussian filter.
You need to specify if the filter should act on all images defined in the DIC input file, or just a selection of images.
The Spatial-Gauss filter needs a specified radius in pixels. (Default = 25 pixels)
Before clicking the Run Filter-button, specify a name for the filtered data set.

The Gaussian filter acts on the nodal displacement in the DIC results, and the nodal displacement $\bm{u}_i = [u_i, v_i]^T$ for a particular node $i$ is given by a weighted sum of the nodal displacements of all $N$ active nodes in the mesh:

$\bm{u}_i = \sum_{j=1}^N{{w_{ij}}\bm{u}_j}$

where the wheight $w_{ij}$ is given by a gaussian bell-curve:

$w_{ij} = ae^{-{|\bm{x}_i - \bm{x}_j|^2}/2r^2}$

The specified radius $r$ is provided by the user, and the constant $a$ is chosen such that $\sum{w_{ij}} = 1$

A large radius $r$ will give a wide bell curve (weighting curve) and thus, a high degree of smoothing, while a small radius will give a narrow bell curve and little smoothing of the nodal displacements.

Time - Moving Average

The moving-average filter acts on each node separately and smooths its displacement in time.

To define a filter, you must first specify the source. By default it is the Raw DIC Data, but it can also be any previosly filtered data.
The filter type is now set as Time - Moving Average.
You need to specify if the filter should act on all images defined in the DIC input file, or just a selection of images.
Specify a moving average degree. This means how many time steps (frame id’s) should be used in calculating a specific average displacement.

A high degree means high degree of smoothing, while a small number means little smoothing of the nodal displacements in time.

Before clicking the Run Filter-button, specify a name for the filtered data set.