Commit 558ceca7 authored by Pierre Paleo's avatar Pierre Paleo
Browse files

Update markdown admonition tags

parent 2e51ab8d
......@@ -99,7 +99,8 @@ API: [MunchDeringer](apidoc/nabu.preproc.rings) and [CudaMunchDeringer](apidoc/n
A flats distortion estimation/correction is available. If activated, each radio is correlated with its corresponding flat, in order to determine and correct the flat distortion.
```warning:: This is currently quite slow (many correlations), and only supported by the "full radios pipeline"
```{warning}
This is currently quite slow (many correlations), and only supported by the "full radios pipeline"
```
Configuration file: `[preproc]`: `flat_distortion_params` and `flat_distortion_correction_enabled`
......@@ -130,7 +131,8 @@ API : [PaganinPhaseRetrieval](apidoc/nabu.preproc.phase) and [CudaPaganinPhaseRe
A single distance CTF retrieval method is also available.
``` warning:: Using this phase retrieval method triggrers the use of a new pipeline, which is less tested that the ones already available. Also, this new pipeline is currently only available with the CPU backend, meaning the overall process will be relatively slor.
```{warning}
Using this phase retrieval method triggrers the use of a new pipeline, which is less tested that the ones already available. Also, this new pipeline is currently only available with the CPU backend, meaning the overall process will be relatively slow.
```
Configuration file: `[phase]`: `method = CTF`, `delta_beta`, `ctf_geometry`, `ctf_translations_file` and `ctf_advanced_params`.
......
......@@ -24,7 +24,8 @@ nabu nabu.conf --slice 500-599
will reconstruct 100 slices: 500, 501, 502, ..., 599 included.
``` warning:: Indexing is zero-based. This means that to reconstruct the first slice, you have to run nabu nabu.conf --slice 0 (or put start_z = 0 in the configuration file)
```{warning}
Indexing is zero-based. This means that to reconstruct the first slice, you have to run nabu nabu.conf --slice 0 (or put start_z = 0 in the configuration file)
```
## Examples
......
......@@ -28,7 +28,8 @@ there is notably a factor of `1/(4*pi**2)` between the conventions.
This method, although based on Contrast Transfer Function (CTF) rather than Transport of Intensity (TIE), can also be expressed as a single step "filtering" processing. See for example [[2]](#references).
``` warning:: This method is not implemented for now.
```{warning}
This method is not implemented for now.
```
## Unsharp masking
......
......@@ -3,8 +3,8 @@
Version 2020.2.0 is a "preview" of a major release foreseen for September 2020. It brings new features since 2020.1.0.
``` note:: The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```{note}
The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```
## Highlights
......@@ -20,7 +20,7 @@ For a quick reconstruction, you can perform binning on the data. There are three
### Automatic center of rotation estimation
The default `rotation_axis_position` value (if left blank) is half of the detector width. The value `rotation_axis_position = auto` can now be provided, so that the value will be automatically estimated.
The default `rotation_axis_position` value (if left blank) is half of the detector width. The value `rotation_axis_position = auto` can now be provided, so that the value will be automatically estimated.
In the current version, automatic estimation lacks robustness and does not work for half tomography.
......
......@@ -3,8 +3,8 @@
Version 2020.3.0 is a major release for ESRF User Service Mode restart.
``` note:: The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```{note}
The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```
## Highlights
......@@ -20,11 +20,11 @@ The volume is automatically divided in sub-volumes. Each sub-volume is reconstru
### Alignment tools
A variety of alignment functions is now available in `nabu.preproc.alignment`.
A variety of alignment functions is now available in `nabu.preproc.alignment`.
### Tiff and JPEG2000 basic support
The default output file format is HDF5. However, the reconstructions can now be saved in tiff and jpeg2000. In this case, there will be one file per reconstructed slice.
The default output file format is HDF5. However, the reconstructions can now be saved in tiff and jpeg2000. In this case, there will be one file per reconstructed slice.
Mind that the current version has the following limitations:
- Tiff: only float32 data is saved (no normalization)
......
......@@ -3,8 +3,8 @@
Version 2020.4.0 adds a number of features, notably command line tools for manipulating datasets.
``` note:: The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```{note}
The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```
## Highlights
......@@ -24,10 +24,10 @@ The paramater is `rotation_axis_position` (in section`[reconstruction]`), the va
### Histograms
Nabu can now compute a histogram of the reconstructed volume. This histogram can serve for further volume processing (conversion to uint16 for example). The computation is done while the data is still in memory to avoid extraneous disk reads/writes.
Nabu can now compute a histogram of the reconstructed volume. This histogram can serve for further volume processing (conversion to uint16 for example). The computation is done while the data is still in memory to avoid extraneous disk reads/writes.
The option is available in a new section `[postproc]` with the parameter `output_histogram = 1`.
A command line tool is also available to merge the histogram of multiple volumes: `nabu-histogram <file1> <file2> ... <output>`.
A command line tool is also available to merge the histogram of multiple volumes: `nabu-histogram <file1> <file2> ... <output>`.
### Use darks/flats from another dataset
......@@ -48,5 +48,6 @@ Projections have a varying "z" identified by the `entry/sample/z_translation` ke
The command line tool `nabu-zsplit` enables to split a file into distinct "z" to reconstruct each volume individually.
``` warning:: This is a temporary solution. This will be natively be handled by nxtomomill in the future.
```
\ No newline at end of file
```{warning}
This is a temporary solution. This will be natively be handled by nxtomomill in the future.
```
......@@ -3,8 +3,8 @@
Version 2020.5.0 is a major version aiming at stabilizing fixes and features added over weeks.
``` note:: The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```{note}
The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```
## Highlights
......@@ -16,7 +16,7 @@ This section highlights some of the available [features](features.md).
In the configuration, nabu now provides four methods for estimating the Center Of Rotation (CoR):
- `centered`: a fast and simple auto-CoR method. It only works when the CoR is not far from the middle of the detector. It does not work for half-tomography.
- `global`: a slow but robust auto-CoR.
- `sliding-window`: semi-automatically find the CoR with a sliding window. You have to specify on which side the CoR is (left, center, right).
- `sliding-window`: semi-automatically find the CoR with a sliding window. You have to specify on which side the CoR is (left, center, right).
- `growing-window` : automatically find the CoR with a sliding-and-growing window. You can tune the option with the parameter 'cor_options'.
Advanced options for each of these methods can be tuned with the `cor_options` parameter.
......@@ -42,8 +42,8 @@ The reason is, sometimes several reconstruction are launched simultaneously on d
### Use of "magnified pixel size"
The HDF5-NX file used as an input contains two pixel sizes: detector pixel size, and "magnified pixel size" accounting for optics effects. Until version `2020.4.1`, the former was taken for scaling the reconstruction and computing the Paganin filter. From `2020.4.1` and on, the magnified pixel size will be used.
The HDF5-NX file used as an input contains two pixel sizes: detector pixel size, and "magnified pixel size" accounting for optics effects. Until version `2020.4.1`, the former was taken for scaling the reconstruction and computing the Paganin filter. From `2020.4.1` and on, the magnified pixel size will be used.
This has mainly two effects:
- For phase retrieval, this change impacts the "blurring effect" but not the gray values scale.
When using a smaller pixel size, the delta/beta value will have to be much smaller to obtain a result similar to the output of previous versions. More precisely, if `magnified_pixel_size = pixel_size/10`, then a result obtained with the parameters `(pixel_size, delta_beta)` using `nabu < 2020.4.1` can be obtained by using parameters `(magnified_pixel_size, delta_beta/100)` in `nabu >= 2020.4.1`.
- For reconstruction, this change impacts the final gray values range, as the reconstructions values are normalized with (horizontal) pixel size.
\ No newline at end of file
- For reconstruction, this change impacts the final gray values range, as the reconstructions values are normalized with (horizontal) pixel size.
......@@ -3,8 +3,8 @@
Version 2021.1.0 is a major version bringing many new features.
``` note:: The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```{note}
The changelog is available at https://gitlab.esrf.fr/tomotools/nabu/-/blob/master/CHANGELOG.md
```
## Highlights
......@@ -13,16 +13,16 @@ This section highlights some of the available [features](features.md).
### Sinogram-based center of rotation estimation
The Center of Rotation (CoR) can now be estimated from a sinogram, instead of two full projections. **This will only work for 360 degrees scans (not 180 degrees).**
To enable CoR estimation from a sinogram, the new parameter `cor_slice` has to be set to a number. This number specifies that the CoR estimation should be done on the sinogram corresponding to this slice number.
The Center of Rotation (CoR) can now be estimated from a sinogram, instead of two full projections. **This will only work for 360 degrees scans (not 180 degrees).**
To enable CoR estimation from a sinogram, the new parameter `cor_slice` has to be set to a number. This number specifies that the CoR estimation should be done on the sinogram corresponding to this slice number.
A new CoR finding method is also available: `rotation_axis_position = sino-coarse-to-fine`. This method only works on (360 deg) sinograms.
### Projections rotation and detector tilt auto-detection
The nabu pipeline can now rotate the projections on the fly, with the parameter `rotate_projections`. It can be useful to correct the effects of a detector tilt, or certain rotation axis tilt in some cases.
The nabu pipeline can now rotate the projections on the fly, with the parameter `rotate_projections`. It can be useful to correct the effects of a detector tilt, or certain rotation axis tilt in some cases.
A new command-line interface tool `nabu-rotate` is also available.
Nabu can now attempt at estimating the detector tilt, if any. If a tilt is detected, then the projection images will be rotated by the found angle value.
Nabu can now attempt at estimating the detector tilt, if any. If a tilt is detected, then the projection images will be rotated by the found angle value.
The parameters in configuration file are `tilt_correction` and `autotilt_options` in `[preproc]`.
### Double flat-field CLI tool
......@@ -31,8 +31,8 @@ The parameters in configuration file are `tilt_correction` and `autotilt_options
### Save and resume from sinograms
It is now possible to save arbitrary processing steps, and to resume the processing pipeline from a saved step. This should dramatically speed up the process of "reconstructing - editing configuration - reconstructing - etc".
In the new section `[pipeline]`, there are two corresponding options `save_steps` and `resume_from_step`.
It is now possible to save arbitrary processing steps, and to resume the processing pipeline from a saved step. This should dramatically speed up the process of "reconstructing - editing configuration - reconstructing - etc".
In the new section `[pipeline]`, there are two corresponding options `save_steps` and `resume_from_step`.
For example, to set a check point to the "sinogram" generation, set `save_steps = sinogram`.
Both `save_steps` and `resume_from_step` can be specified at the same time. If the file is not found, then all the processing is done in normal mode, and the steps are saved to the file. Otherwise, the pipeline resumes from the step. The file where data is dumped can be specified with the parameter `steps_file`.
......@@ -43,12 +43,12 @@ A new rings artefacts removal technique has been added. It is the implementation
### CTF phase retrieval
A new single-distance phase retrieval method is available: CTF. If used, it triggers the use of a new processing pipeline (more details below).
A new single-distance phase retrieval method is available: CTF. If used, it triggers the use of a new processing pipeline (more details below).
Additional parameters in configuration file are `ctf_geometry`, `ctf_translations_file` and `ctf_advanced_params` in `[phase]`.
### Flats distortion correction
Added support for flats distortion correction. If activated, each radio is correlated with its corresponding flat, in order to determine and correct the flat distortion.
Added support for flats distortion correction. If activated, each radio is correlated with its corresponding flat, in order to determine and correct the flat distortion.
This is currently quite slow (many correlations).
The parameters in configuration file are `flat_distortion_params` and `flat_distortion_correction_enabled` in `[preproc]`.
......@@ -56,7 +56,7 @@ The parameters in configuration file are `flat_distortion_params` and `flat_dist
### New processing pipeline (CPU only !)
In some cases, the processing has to be done on the whole projection image, rather than a part of it. For example, the CTF filter spans the entire image support. This contrast with the approach used by nabu up to now : radios were processed by parts, and then the results were merged.
In some cases, the processing has to be done on the whole projection image, rather than a part of it. For example, the CTF filter spans the entire image support. This contrast with the approach used by nabu up to now : radios were processed by parts, and then the results were merged.
Nabu now offers a new pipeline which can process "full images", enabling the use of CTF and images rotation. Mind that for now, **only the CPU backend of this pipeline is available**, so it might be slow compared to the existing pipeline.
### EDF output format
......
# Configuration validator
``` note:: This page is intended for Nabu developers.
```{note}
This page is intended for Nabu developers.
```
## Principle
......@@ -45,9 +46,10 @@ A dataset is either
From the dataset, many parameters are extracted and put in a dictionary.
``` important:: All the parameters that can be retrieved from the dataset should be avoided in the configuration file. The purpose of the "Dataset Analyser" component is to avoid filling too much the configuration file. For example, the number of pixels in each dimension is not relevant in the configuration file.
```{important}
All the parameters that can be retrieved from the dataset should be avoided in the configuration file. The purpose of the "Dataset Analyser" component is to avoid filling too much the configuration file. For example, the number of pixels in each dimension is not relevant in the configuration file.
```
## Validator, second stage
At this point, we have two dictionaries of parameters: one from the parsing/validation of the configuration file, and one from the dataset analyzer.
\ No newline at end of file
At this point, we have two dictionaries of parameters: one from the parsing/validation of the configuration file, and one from the dataset analyzer.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment