Navigation: Utilities >
Spatial Filter Utility
|
Navigation: Utilities > Spatial Filter Utility |
|
The Spatial Filter Utility dialog is used to apply matrix filters to surface data. There are six spatial filters available: Median, Mean, Maximum, Minimum, Gaussian, and Laplacian. These filters are typically used to improve images or search for details within an image. The available filter sizes are 3x3, 5x5, 7x7, and 9x9 samples. Please note that the filtered surface will be reduced in size depending on the size of the filter matrix applied. The residual surface is simply the filtered surface subtracted from the original surface; therefore, it will contain the surface detail removed by the filtering process.
|
The median filter can useful in remove noise (small spikes or drop-outs) from a surface. The median filter arranges the data points in the region of the matrix in least to greatest order. The middle value is then used as the filtered sample value.
Note how a 3x3 median filter applied to surface to the right removed the small spikes on the surface.
|
|
|
|
The mean filter is used to smooth the surface. The samples within the region of the matrix are averaged and this average is used as the filtered sample value. The larger the filter matrix size, the larger the smoothing effect.
Note how a 9x9 mean filter applied to surface to the right smoothed the overall appearance of the surface.
|
|
|
|
The maximum filter selects the sample with the maximum value within the matrix region and uses that as the new sample value.
This residual surface created by this filter is often used to find outliers on the surface.
|
||
|
The minimum filter selects the sample with the minimum value within the matrix region and uses that as the new sample value.
This residual surface created by this filter is often used to find outliers on the surface.
|
||
|
The Gaussian matrix filter is similar to the mean filter except that it favors the lowest point in the region of the matrix.
Note that this filter type is only available as a 3x3 matrix.
Matrix kernel:
1 2 1 2 4 2 1 2 1
|
||
|
The Laplacian filter is able to detect sharp transitions in the surface data.
Note how the filtered surface has a very obvious spike at the location of the outlier in the original surface.
Matrix kernel:
-1 -1 -1 -1 9 -1 -1 -1 -1
|
|
|
