genre | Kernel |
author | Sven Hansen |
package | MeVisLab/Standard |
dll | MLKernel |
definition | MLKernel.def |
see also | ExtendedConvolution, Rank, KernelEditor, Morphology, BitMorphology, KernelExample, Sigma, LoG |
keywords | enhance, kernel, input, filter, laplace, gauss, sobel, average, smooth, edge, interval, sphere, spherical, structure, element |
The module Convolution implements generic kernel-based filtering (linear and shift-invariant).
The implemented kernel functions are defined by
where g denotes the output, f the input, t the kernel and k is a 6D multi-index.
Please note that in this module not a point spread function (impulse response) is supplied by the user but a kernel function. The point spread function h is related to the kernel via mirroring (origin is the kernel center):
One can use some simple pre-defined kernels like Average, Gauss, Laplace and Sobel, or supply a custom defined kernel via the kernel-editor module.
Optionally, a threshold interval can be used to select a subset of voxels of the input image to which the filtering is applied. By using this threshold interval the filtering can be applied selectively to regions in the image.
Border Handling: Enum | Use (useIntervalOfFilteredVoxels): Bool |
External Kernel: String | Use (predefKernel): Enum |
Fill Value: Double | |
Kernel Output: String | |
kernelInfo: String | |
Max: Double | |
Min: Double | |
referenceExtentMode: Enum |
Defines the border handling mode.
See MLGuide/BorderHandling for details.
Values:
Title | Name |
---|---|
No Pad | NoPad |
Pad Src Fill | PadSrcFill |
Pad Dst Fill | PadDstFill |
Pad Dst Fill With Orig | PadDstFillWithOrig |
Pad Src Undefined | PadSrcUndefined |
Pad Dst Undefined | PadDstUndefined |
Pad Src Clamp | PadSrcClamp |
Sets the fill value for certain Border Handling modes.
Sets the minimum value threshold for outputting a subset of voxel.
Sets the maximum value threshold for outputting a subset of voxel.
If checked, the module outputs only on a subset of voxels defined by a value range.
Sets an external kernel as a string.
Use the module KernelEditor for editing the kernel or set the kernel by scripting.
Shows the used kernel as a string for connecting to other modules that use an external kernel.
Defines the applied filter.
Values:
Title | Name | Description |
---|---|---|
3x3 Average Kernel | 3x3 Average Kernel | All Elements of this kernel have normalized and equal coefficents. 1/9 1/9 1/9
1/9 1/9 1/9
1/9 1/9 1/9
|
5x5 Average Kernel | 5x5 Average Kernel | All Elements of this kernel have normalized and equal coefficents. 1/25 1/25 1/25 1/25 1/25
1/25 1/25 1/25 1/25 1/25
1/25 1/25 1/25 1/25 1/25
1/25 1/25 1/25 1/25 1/25
1/25 1/25 1/25 1/25 1/25
|
7x7 Average Kernel | 7x7 Average Kernel | All Elements of this kernel have normalized and equal coefficents. |
10x10 Average Kernel | 10x10 Average Kernel | All Elements of this kernel have normalized and equal coefficents. |
25x25 Average Kernel | 25x25 Average Kernel | All Elements of this kernel have normalized and equal coefficents. |
3x3 Gauss Kernel | 3x3 Gauss Kernel | Applies a 3x3 Gaussian kernel. Smoothes image and emphasizes the inner voxels. 1/16 2/16 1/16
2/16 4/16 2/16
1/16 2/16 1/16
|
5x5 Gauss Kernel | 5x5 Gauss Kernel | Applies a 5x5 Gaussian kernel. |
Laplace Kernel | Laplace Kernel | Applies a 3x3 Laplacian kernel. 0 1 0
1 -4 1
0 1 0
|
Laplace2 Kernel | Laplace2 Kernel | Applies a 3x3 Laplacian kernel for edge detection. This one is taking the second derivation into account. 1 2 1
2 -12 2
1 2 1
|
Sobel1 Kernel | | Sobel1 Kernel | | Applies a 3x3 Sobel1 kernel. Used for edge detection in vertical direction and smoothing in horizontal direction. 1 0 -1
2 0 -2
1 0 -1
|
Sobel2 Kernel - | Sobel2 Kernel - | Applies a 3x3 Sobel2 kernel. Used for edge detection in horizontal direction and smoothing in vertical direction. 1 2 1
0 0 0
-1 -2 -1
|
Sobel3 Kernel | Sobel3 Kernel | Applies a 3x3 Sobel3 kernel. Used for edge detection in diagonal direction and smoothing in contrary direction. 0 -1 -2
1 0 -1
2 1 0
|
Sobel4 Kernel / | Sobel4 Kernel / | Applies a 3x3 Sobel4 kernel. Used for edge detection in diagonal direction and smoothing in contrary direction. -2 -1 0
-1 0 1
0 1 2
|
Id Kernel | Id Kernel | Just copies the input image to the output. |
From Kernel Input | From Kernel Input | Uses the kernel specified in the kernel input field. Usually this will be a kernel passed via a field connection from the text field of a kernel editor. Note that a kernel defined once can be saved/loaded with the operator state normally even if it is not visible as kernel matrix. Please note that the kernel editor specifies a kernel function, not a point spread function. |