Filtering
Filters provide an easy to use, yet powerful tool to remove interfering elements (like noise or ice gradients) from an image. Here, a set of different filters are provided which all apply a filter onto the input stack of images.
Usage
A stack of images/3Ds or an individual 3D will be subjected to the selected filter with its respective filter parameters. For more detail on the capabilities of the available filters and their parameters, consult the respective process section below. Since most of the filters just have one input (e.g. stack of images) and one output (e.g. stack of images w/ applied filter), only divergent in-/outputs are mentioned.
Processes
Butterworth
Applies a Butterworth Filter with the chosen parameters to the images. A butterworth filter has a smooth transfer function without any discontinuity. The filter functions used in this logic are:
B(Low Pass)= 1/ (1+[D(u,v)/DL ]2n)
B(High Pass)= 1- (1/ (1+[D(u,v)/DH]2n))
where D(u,v) is the distance from the origin, n is the order of the filter and DL/DH is the cut frequency.
| Parameters | Description |
|---|---|
| Type | Select whether you want to apply a highpass or lowpass filter |
| pass-band | Order of the filter n |
| showFilterFunction | Use this option to change the output consisting of complex images showing the filter function instead of the filtered images |
| stop-band | Upper or lower frequency cut-off DL / DH, depending on whether high- or lowpassfilter is selected |
Convolution
| Parameters | Description |
|---|---|
| Flip | Description of this parameter |
| Scale factor | more description |
Cosine BP/LP/HP
This processor applies a cosine bandpass, lowpass or highpass filter, respectively, on the images.
| Parameters | Description |
|---|---|
| Cosine width [0…1] | This value sets the width of the cosine function in the filter mask. |
| Upper frequency [0…1] | This value sets the lowpass frequency used for the filtering mask. |
| Lower frequency [0…1] | This value sets the highpass frequency used for the filtering mask. |
| showFilterfunction | If enabled, the output shows the filter mask according to the set parameters instead of the filtered images. |
Difference of gaussians
The DoG is a wavelet mother function of null total sum which approximates the Mexican Hat Wavelet by subtracting a wide from a narrow Gaussian. This filter is applied on each input image and results in feature enhancement through edge detection. Further details can be found here and here . 
| Parameters | Description |
|---|---|
| Narrow standard deviation | This value sets narrower standard deviation for the DoG method. It describes one of the kernel functions used to perform the filtering procedure. If Narrow standard deviation is larger than Wide standard deviation, the values are swapped. |
| Wide standard deviation | This value sets wider standard deviation for the DoG method. It describes one of the kernel functions used to perform the filtering procedure. If Narrow standard deviation is larger than Wide standard deviation, the values are swapped. |
Flat histogram
| Parameters | Description |
|---|---|
| Some changeable parameter | Description of this parameter |
| → and its sub-parameter | more description |
| Next main parameter | and more more more |
| → and its sub-parameter | … descriptions |
Gauss BP/LP/HP
This processor applies a Gaussian bandpass, lowpass or highpass filter, respectively, on the images.
| Parameters | Description |
|---|---|
| Upper frequency [0…1] | This value sets the lowpass frequency used for the filtering mask. |
| Lower frequency [0…1] | This value sets the highpass frequency used for the filtering mask. |
| Show filter function | Gives the filter function as output instead of the filtered images. |
| Transmission | Defines the width of the cutoff region of the Gaussian function. |
Gaussian bilateral filtering
This processor applies a Gaussian bilateral filter on image data.
| Parameters | Description |
|---|---|
| Range kernel std dev | This value sets the standard deviation of the range kernel used to smooth differences in intensities (Gaussian function). |
| Spatial kernel std dev | This value sets the standard deviation of the spatial kernel used to smooth differences in intensities (Gaussian function). |
| Size of window | This value sets the size of the window used to compute the intensity values during the filtering process. |
Hann window filter
| Parameters | Description |
|---|---|
| Hann window factor | Description of this parameter |
| Show filter function | Gives the filter function as output instead of the filtered images. |
Kernel convolution
Sobel
This processor performs edge detection, resulting in a binary image (part of edge =1 , rest =0 ). This processor is based on the Sobel Operator. More information can be found here.