EM Processing
This group lists image processing functions that are designed for processing FIB-SEM images. Some of the functions can also be found in other groups of image processing functions. This group is only visible if you have the license for EM Processing Toolbox.
See also
Change Pixel Type
This method allows you to change the pixel type of an image. This can be useful if you want to compare or combine images that have different pixel types.
|
Parameter |
Description |
|
|---|---|---|
|
Pixel Format |
Select the desired pixel format from the dropdown list. |
|
|
- |
8 Bit B/W |
The output image is a monochrome image, the whole-number gray values of which can lie in the range from 0 to 255. |
|
- |
16 Bit B/W |
The output image is a monochrome image, the whole-number gray values of which can lie in the range from 0 to 65535. |
|
- |
32 Bit B/W Float |
The output image is a monochrome image with real numbers as pixel values. |
|
- |
2x32 Bit Complex |
The output image is a monochrome image with complex numbers (real part and imaginary part) as pixel values. Such images are generally created by means of transformation into the Fourier space. |
|
- |
24 Bit RGB |
The output image is a color image, the whole-number color values of which in the red, green, and blue channels can lie in the range from 0 to 255. |
|
- |
48 Bit RGB |
The output image is a color image, the whole-number color values of which in the red, green, and blue channels can lie in the range from 0 to 65535. |
|
- |
2x32 Bit RGB Float |
The output image is a color image with real numbers as color values in the red, green and blue channels. |
|
- |
3x64 Bit RGB Complex |
The output image is a color image with complex numbers (real part and imaginary part) in the red, green and blue channels. Such images are generally created by means of transformation into the Fourier space. |
Create Image Subset
Image Analysis Results
Note that if your image contains analysis results, the analysis results are deleted when you execute this function.
Lightfield 4D Raw Data
If you want to use the function with unprocessed Lightfield 4D raw data, you must not extract single phases but have to extract all phases. Lightfield 4D processing always needs the data of all phases (lenses). Additionally, if you want to extract a region, Lightfield 4D raw data images need a minimal region size. A message is displayed in the tool in both situations to indicate the appropriate setting.
This method allows you to extract parts from one image and use these to create a new image. You can select these parts freely from the individual dimensions of the image. Each of the parameter sections is only visible if the corresponding dimension is present in the input image.
|
Parameter |
Description |
|
|---|---|---|
|
Channels |
Selects which channels of the input image are used. All channels are selected by default. To deselect a channel, click on the respective channel button. |
|
|
Z-Position, Time, Block, Scene, Phase |
Here you can select which parts of the input image you want to use for the resulting image. |
|
|
- |
Extract All |
If selected, all parts of the corresponding image are extracted. |
|
- |
Extract Single |
If selected, you can select a single image to be extracted. |
|
- |
Extract Range |
If selected, you can select a certain range of images to be extracted. |
|
- |
Extract Multiple |
If selected, you can select several continuous ranges and individual sections. Enter one or more sections that you want to select in the input field. To do this, enter the first section, followed by a minus sign, and then the last section. If you want to define an interval, after the last section enter a colon and then the interval. The entry "2-10:2" means that every second section is selected from section 2 to section 10. Enter a comma after the first section if you want to define another section. You can also select individual sections separated by commas. By entering "2-10:2,14-18,20,23", you select every second section from section 2 to section 10, followed by sections 14 to 18, as well as sections 20 and 23. |
|
- |
Get current position |
Adopts the position from the current display in the image area. |
|
- |
Interval |
Activated: Interval mode is active. The Interval spin box/input field appears. Enter the desired interval here. E.g. if you enter the value 2 only every 2nd value from the range is considered. |
|
Region |
Here you can select if you want to use the entire image or just a region (ROI) of the input image. |
|
|
- |
Full |
Select this option to use the full image for the new image. |
|
- |
Rectangle region (ROI) |
Select this option to draw in a rectangle region of interest, which will be used for creating a new image. If a rectangle region was drawn in, you can see and change its coordinates by editing the X/Y/W/H input fields. |
|
- |
Keep tiles |
Only has an effect, if a region (ROI) is defined. Activated: Extracts the drawn in region including the complete tiles. This setting is recommended when you want to apply DCV processing functions on the resulting image. |
Denoise
This method removes noise from images using wavelet transformations or total variation. The process of denoising an image with wavelet transformations can be broken down into the following three parts:
- Calculate the wavelet transform of the noisy image.
The wavelet transformation can be calculated by the method Real Wavelets and Complex Wavelets. - Modify the noisy wavelet coefficients.
This is done by using bivariate shrinkage with local variance estimation (thresholding). [Bivariate Shrinkage with Local Variance Estimator, Levent Sendur and Ivan W. Selesnick, IEEE Signal Processing Letters, Vol. 9, No. 12, December 2002] - Compute the inverse transform using the threshold coefficients.
|
Parameter |
Description |
|
|---|---|---|
|
Method |
||
|
- |
Complex Wavelets |
The Dual Tree Complex Wavelet transform provides better results due to the fact that it is nearly direction invariant and makes more directional sub bands available. The results will be less prone to block-artefacts. However, this method is computationally more intense and therefore takes longer. |
|
- |
Real Wavelets |
The real wavelet transform only considers three sides (XYZ) and is therefore faster. However, the result can show block artefacts. |
|
- |
Total Variation |
An algorithm based on A. Chambolle, "An Algorithm for Total Variation Minimization and Applications", J. Math. Imaging and Vision 20 (1-2): 89-97, 2004. |
|
Strength |
Here you adjust the strength with which the function is applied. |
|
See also
Enhance Local Contrast
With this function you can improve the contrast of your image.
|
Parameter |
Description |
|---|---|
|
Clip Limit (%) |
Defines the value at which the histogram is clipped. It is used to avoid oversaturation of the image in homogeneous areas. It is determined by the normalized histogram of the local region. The higher the value is set, the higher the contrast will be. |
|
Region Size (%) |
Defines the size of every single region or tile where the local histogram is calculated as percentage of the image size. For example, a region size of 50% has the size of half the image. |
|
Bins |
Defines the number of histogram bins that is used to create the contrast enhancing transformation. A higher number of bins covers a greater dynamic range but increases calculation time. |
|
Process Tiles Separately |
Only visible for tile images. Activated: Processes each tile region separately. |
Difference between Result and Preview
The actual result of the function and the preview you see in the preview window can be different. This is caused by the parameter Region Size (%) and the different size of the preview and original image. The function splits the image into small regions and processes them individually based on the histogram of every single region. The histogram is usually related to the absolute region size. For the preview only the visible part of the image in the view is processed.
Gauss
This method allows you to reduce noise in an image. Each pixel is replaced by a weighted average of its neighbors. The neighboring pixels are weighted in accordance with a two-dimensional Gauss bell curve.
|
Parameter |
Description |
|---|---|
|
Sigma |
Here you can adjust the sigma value. If the Show All mode is activated, you can adjust the values in each dimension individually. |
Highpass
This method performs high-pass filtering. The high pass filter is defined as the difference between the original image and the low-pass filtered original.
|
Parameter |
Description |
|
|---|---|---|
|
Normalization |
Defines how out-of-range pixel values are mapped. The calculated pixel values of the output image may be out-of-range and are mapped into the available range. |
|
|
- |
Clip |
Values exceeding the pixel value range are set to the highest value available (white), values falling short of the pixel value range are set to the lowest value available (black). The effect corresponds to underexposure or overexposure. This means that in some cases information is lost.
|
|
- |
Automatic |
Normalizes the pixel values automatically to the available pixel value range. The highest resulting value is mapped to the maximum pixel value, the lowest resulting value to 0. As a result, the whole range of resulting pixel values is compressed evenly.
|
|
- |
Wrap |
If a resulting value is larger than the maximum pixel value of the image, the difference exceeding the maximum pixel value is added to 0. Similarly, if a resulting value is below 0, the resulting pixel value is the maximum pixel value minus the difference falling below 0.
|
|
- |
Shift |
Normalizes the output to the value "pixel value + maximum pixel value/2". As a result, all resulting values are mapped to the available value range. The middle value of the pixel value range remains constant. Values left and right of the middle value are changed progressively, so that values inside the pixel value range are changed only slightly. Values outside the pixel value range are changed strongly and mapped to the fringes of the pixel value range.
|
|
- |
Absolute |
Converts negative pixel values into positive values. Positive pixel values exceeding the maximum pixel value are set to the maximum pixel value.
|
|
Parameter |
Description |
|
|---|---|---|
|
Count |
Here you set the number of repetitions. I.e. the number of times the function is applied sequentially to the respective result of the filtering. The effect is increased correspondingly. |
|
|
Kernel Size |
You can set the filter size in the x-, y- and z-direction, symmetrically around the subject pixel. This should be the size of the transition region between objects and background match. |
|
Import SmartFIB TIFFs
This function allows you to import SmartFIB stacks of Crossbeam microscopes into ZEN.
|
Parameter |
Description |
|---|---|
|
Select Files |
Opens a file browser to select the images for import. |
|
Z-Spacing |
Sets the slice distance in z. Per default, Auto is activated, and the distance is calculated automatically with information from the image metadata of the first and last selected image. |
|
XY-Scaling |
Sets the scaling in x and y. Per default, Auto is activated, and the scaling is automatically taken from the image metadata of the first selected image. |
|
Sample Angle |
Sets the angle of the sample. Per default, Auto is activated, and the sample angle is automatically calculated from the image metadata or set to the default of 54 degree and the image is rendered with a 90 degree tilt (if no information is available in the metadata). In this case the SmartFIB stack will be displayed perpendicular to the images acquired on a light microscope. |
|
Read XY Offsets |
Activated: Reads the xy offset of the individual slices from the metadata. Note that this can lead to a slanted z-stack depending on the sample orientation during imaging and the metadata. |
Median
This method allows you to reduce noise in an image. Each pixel is replaced by the median of its neighbors. The size of the area of the neighboring pixels considered is defined by a quadratic filter matrix. The modified pixel is the central pixel of the filter matrix. The median is the middle value of the gray values of the pixel and its neighbors sorted in ascending order.
|
Parameter |
Description |
|---|---|
|
Kernel Size |
Here you can adjust the size of the filter matrix. If the Show All mode is activated, you can adjust the values in X, Y and Z direction individually. |
Not
This function performs a binary "not" operation on all bits of the binary representation of an input pixel's gray value. A 0-bit in the input pixel results in an 1-bit in the corresponding output pixel, an 1-bit in the input gets a 0-bit in the output. For integral image types, the resulting output gray value is the difference of the maximum possible gray value minus the input gray value, but for float image type the results are strange due to the inhomogeneous float format.
Remove Stripes
This function allows you to remove stripe artifacts from your image. It is recommended to use the output preview to determine the suitable values for your parameters. You can display the output preview window by clicking on Preview in the Output tool.
|
Parameter |
Description |
|
|---|---|---|
|
Method |
Selects the method you want to use for the removing the striping artifacts in your image. |
|
|
– |
GPU VSNR |
Only available if you have a NVIDIA GPU in your machine. |
|
– |
GPU |
Only available if you have a NVIDIA GPU in your machine. |
|
– |
CPU |
Uses a CPU optimized algorithm to remove artifacts from the image. |
|
Iterations |
Sets the number of iterations after which the calculation is ended. |
|
|
First Filter |
Displays the settings of the first filter to remove vertical or horizontal stripes in your image. |
|
|
– |
Noise Level |
Sets the intensity level of the stripes in your image. |
|
– |
Width |
The calculation for the removal of stripes in the image is based on a Gaussian curve. This parameter sets the width of the curve. |
|
– |
Height |
The calculation for the removal of stripes in the image is based on a Gaussian curve. This parameter sets the elongation of the curve. |
|
– |
Angle |
Sets the angle for stripes in your image. An angle of 0 corresponds to vertical stripes. An angle of 90 (or -90) corresponds to horizontal stripes. |
|
Second Filter |
To remove a second set of stripes (e.g. if you have both vertical and horizontal stripes in your image), you can activate a second filter. For information about the parameters, see the descriptions above. |
|
|
– |
Enabled |
Activated: Displays the settings for a second filter to remove striping artifacts in the image. |
Shading Correction
This method allows you to improve images in which the quality has been impaired by uneven illumination or vignetting.
If you want to perform shading correction before an experiment (recommended), you have to use the shading correction function in the Camera tool, in the Post Processing Section.
|
Parameter |
Description |
|
|---|---|---|
|
Shading Mode |
Selects which shading correction should be applied. |
|
|
- |
Channel Specific |
For multichannel references, the information from each channel is applied individually to the input image. The number and type of channels must be the same between input image and shading reference image. This is the correct setting when using a reference image which was created by the Shading Reference for Processing function (e.g. for Axioscan images). |
|
- |
Averaged |
For multichannel references, an average over all channels is calculated and then applied equally to each tile of each channel. This should ensure that a good result can still be achieved with an unequal number of channels or different channel characteristics. |
|
- |
Tile Specific |
In case a shading reference has more than one tile, the tiles of the reference and the tiles of the image that should be corrected are mapped to each other. The tiles must be identical in number and XY position. This can be the case, for example, when using an Axiozoom where shading is primarily determined by the sample mounting situation rather than the optical system. |
|
Automatic |
Activated: The function automatically calculates the reference image for shading correction from the input image. |
|
|
Shading Reference |
Only available in Batch mode and if Global Shading is selected, or Camera Shading is selected and Automatic is deactivated. Enables you to select a reference image for the shading correction. |
|
|
Display Mode |
||
|
- |
Additive |
In this mode the normalized reference image is subtracted from each camera frame. This influences the brightness of the image. |
|
- |
Multiplicative |
In this mode each camera frame is divided by the normalized reference image. This influences the contrast of the image. This is the default setting. |
|
Offset |
Adjust the gray value that will be added on to the newly calculated gray values using the slider or input field. If this results in negative values, these are set to 0. Values that exceed the maximum gray value are set to the maximum gray value. |
|
See also
Slices Replacement
With this function you can replace one or more z-slices of a stack with the respective previous or following slice.
|
Parameter |
Description |
|
|---|---|---|
|
Replace with Next |
Replaces the currently selected z-slice with the following one. |
|
|
Replace with previous |
Replaces the currently selected z-slice with the previous one. |
|
|
Undo |
Reverses the last replacement(s). |
|
|
Redo |
Restores the last undone change(s) for the replacement. |
|
|
Replacement Table |
Only visible if at least one replacement has been set. Displays information about the slice replacement and which Z is replaced. |
|
|
– |
Z |
Displays the index of the z-slice that gets replaced. |
|
– |
Source |
Displays the index of the z-slice which replaces Z. |
|
– |
Replacement Type |
Displays the type of replacement (Next or Previous). |
See also
Sort SmartFIB Tiffs
This function allows you to select a folder with FIB-SEM tiffs and sort these files in separate subfolders according to channel name, number of pixels, image size, and spacing of the images, corresponding to slice thickness. The files are sorted into subfolders of the output folder according to their metadata. This means that each time one of the above mentioned metadata changes from one image to the subsequent image, a new sub folder is created. Thus, subfolders with homogeneous metadata with respect to xyz-scaling and image size are created.
Note that this function only works if the tiff files have their default names assigned by the microscope (e.g. channel0_slice_0001.tiff or slice_0001.tiff)! All tiff files with other names are ignored by the function.
|
Parameter |
Description |
|
|---|---|---|
|
Input Folder |
Selects the folder where the tiff files which should be sorted are saved. Click on |
|
|
Output Folder |
Selects the folder where the subfolders with the sorted tiff files are placed. Click on |
|
|
File Operation |
||
|
– |
Copy |
Copies the images into the output folder. Your data is duplicated and remains unchanged in the input folder. |
|
– |
Move |
Moves the images from your input folder into the output directory. |
|
– |
Hard Link |
Creates links in the output folder to each image file in the input folder, i.e. the space occupied by the images on the hard disk remains unchanged. |
to open a file browser and navigate to the respective folder. Note that this function does not include images from subfolders, all images that you want to be sorted have to be in this input folder!Z-Stack Alignment with ROI
This function allows you to automatically align the individual planes of a z-stack image if they are not positioned precisely above each other.
|
Parameter |
Description |
|
|---|---|---|
|
Quality |
Selects the quality level that you want the function to work with. The calculation of the alignment is based on a so-called image pyramid. The higher the selected quality, the more levels of the image pyramid are used to calculate the alignment and the more precise the alignment will be. However, the higher the selected quality is, the slower the calculation of the alignment will get. |
|
|
– |
Low |
This is the most imprecise but also the fastest calculation of the alignment. It uses a low number of levels (2) of the image pyramid for the alignment calculation. |
|
– |
Medium |
This is a more precise but also a slower calculation of the alignment than the one before. It uses a medium number of levels (3) of the image pyramid for the alignment calculation. |
|
– |
High |
This is a more precise but also a slower calculation of the alignment than the two before. It uses a high number of levels (4) of the image pyramid for the alignment calculation. |
|
– |
Highest |
This is the most precise but also the slowest calculation of the alignment. It uses the highest number of levels of the image pyramid for the alignment calculation. |
|
Registration Method |
Selects the method which is used to align the images. |
|
|
– |
Translation |
The neighboring sections of the z-stack are shifted in relation to each other in the X and Y direction. |
|
– |
Rotation |
The neighboring sections of the z-stack are rotated in relation to each other. |
|
– |
Translation + Rotation |
The neighboring sections of the z-stack are translated and rotated in relation to each other. |
|
Interpolation |
Selects how the interpolation is performed if a pixel is calculated from several individual pixels. |
|
|
- |
Nearest Neighbor |
The output pixel is given the gray value of the input pixel that is closest to it. |
|
- |
Linear |
The output pixel is given the gray value resulting from the linear combination of the input pixels closest to it. |
|
- |
Cubic |
The output pixel is given the gray value resulting from a polynomial function of the input pixels closest to it. |
|
Region |
Selects which parts of the image should be considered for the calculation of the transformation matrix. |
|
|
- |
Full |
Considers the entire image for alignment. |
|
- |
Rectangle Region |
Allows you to draw a region of interest into the image. Only the image information of this region is then considered for the calculation of the transformation matrix for alignment. The resulting transformation matrix will be applied to the full image. |
|
Crop Output |
Not visible if the Third Dimension (if available) is set to a particular dimension instead of 2D Slices. |
|
|
Single Component |
Activated: Displays the image channels. Selects one of the image channels which will be used to calculate the alignment transformation matrix, which then is also applied to the other channel(s). |
|
See also
Z-Stack Equalization
This function equalizes the gray values throughout an entire image stack. It scales the intensities of all slices to a common mean intensity value and standard deviation. The goal is a harmonization of the gray value distribution within a z-stack. With this process, brighter image slices will get darker and darker slices will get brighter.