Commit 7a451577 authored by Martin Schorb's avatar Martin Schorb
Browse files

doc - tifstack importer and other updates

parent 260f6cc0
Pipeline #34841 passed with stage
in 59 seconds
docs/img/webui_convert.png

85.8 KB | W: | H:

docs/img/webui_convert.png

97.7 KB | W: | H:

docs/img/webui_convert.png
docs/img/webui_convert.png
docs/img/webui_convert.png
docs/img/webui_convert.png
  • 2-up
  • Swipe
  • Onion skin
...@@ -38,6 +38,12 @@ This page contains the following elements: ...@@ -38,6 +38,12 @@ This page contains the following elements:
- **dataset metadata file:** the path of the `idoc` file for the SerialEM (Super)Montage. - **dataset metadata file:** the path of the `idoc` file for the SerialEM (Super)Montage.
- **browse:** use this dropdown to browse the directory. To move up (`..`) multiple times, you have to close the selector (`x` on the very right) for each additional step up. - **browse:** use this dropdown to browse the directory. To move up (`..`) multiple times, you have to close the selector (`x` on the very right) for each additional step up.
### Image stack - FIB-SEM
- **dataset root directory:** the directory path where the images are stored. They should be in aplphabetical order according to the layers.
- **Resolution values:** the pixel sizes in `x/y` and the section spacing in `z` in nm.
## Generate Tile Pairs ## Generate Tile Pairs
This step will tell Render which of the tiles are neighbours in `x` and `y` and also in `z`. It will then have a collection of pairs that it can try to match with each other all in parallel. This step will tell Render which of the tiles are neighbours in `x` and `y` and also in `z`. It will then have a collection of pairs that it can try to match with each other all in parallel.
...@@ -84,12 +90,17 @@ and ...@@ -84,12 +90,17 @@ and
There are also two **Tile Viewers** where you can browse the stack and identify two neighbouring tiles to be used for the match trial. There are also two **Tile Viewers** where you can browse the stack and identify two neighbouring tiles to be used for the match trial.
If you click **Explore MatchTrial**, you will get to the Match Trial web interface. It will show the results of the template trial (the images might not be available). If you click **Explore MatchTrial**, the parameters will be used to match the two example tiles and the Match Trial web interface will open in a new browser tab. It will show the results of the trial using the template parameters and the selected tiles.
![matchtrial_template](img/match_trial_create.png "Render WebUI matchTrial") If you are happy with the result, you can continue launching the procedure for all tiles. The TrialID will be automatically filled from this trial.
#### if you need to adjust parameters because the matches are not good:
Click **Create New Trial** at the top of the page.
![matchtrial_template](img/match_trial_create.png "Render WebUI matchTrial")
Click **Create New Trial** at the top of the page. This will open the Parameter editor with the parameters defined in the template. This will open the Parameter editor with the parameters defined in the template.
![matchtrial_params](img/match_trial_params.png "Render WebUI matchTrial Parameters") ![matchtrial_params](img/match_trial_params.png "Render WebUI matchTrial Parameters")
...@@ -122,14 +133,16 @@ If you are happy with the parameters, copy the long ID at the very end of the we ...@@ -122,14 +133,16 @@ If you are happy with the parameters, copy the long ID at the very end of the we
There, ***paste it into the text box*** after **Use this Match Trial as compute parameters:**. There, ***paste it into the text box*** after **Use this Match Trial as compute parameters:**.
These parameters will then be used for the search of all the tiles. #### Compute the matches for all tilepairs
The parameters you just defined for the template tile pair will be used for all the tiles in the stack.
With the known number of tile pairs and the estimated run time for each pair from the trial, the necessary compute resources for the stack are predicted and can be checked in **Compute settings**. With the known number of tile pairs and the estimated run time for each pair from the trial, the necessary compute resources for the stack are predicted and can be checked in **Compute settings**.
![comp_set](img/webui_comp_set.png "VolumeAlign WebUI pointmatch compute settings") ![comp_set](img/webui_comp_set.png "VolumeAlign WebUI pointmatch compute settings")
Launching the computation will request the selected resources on the cluster and then launch a Spark instance on the allocated compute nodes that distributes and manages the parallel computation of the point matches. Launching the computation will request the selected resources on the cluster and then launch a Spark instance on the allocated compute nodes that distributes and manages the parallel computation of the point matches.
You will receive an email once the computation is done. It will tell you that the computation was `CANCELLED` but this only means that the resource allocation has been ended after successful computation of the matches. If you get a message referring to a `TIMEOUT`, you have to re-run the computation with more generous resource settings. You will receive an email once the computation starts (after allocation of the resources) and when it is done. It will tell you that the computation was `COMPLETED` if successful and give you error descriptions otherwise. If you get a message referring to a `TIMEOUT`, you have to re-run the computation with more generous resource settings.
## Solve Positions ## Solve Positions
...@@ -163,12 +176,14 @@ You can export a Render stack at any time to disk ("materialize"). ...@@ -163,12 +176,14 @@ You can export a Render stack at any time to disk ("materialize").
![export](img/webui_export_slice.png "VolumeAlign WebUI export") ![export](img/webui_export_slice.png "VolumeAlign WebUI export")
You can **explore slice** to determine the **volume region to consider**. You can **explore slice** to determine the **volume region to consider**. Just drag the rectangle on top of the image. If you click **Zoom In**, the viewer and region boundary will adapt to the selection. **Reset Zoom** will bring the viewer back to displaying ther entrire setcion.
At the moment, `N5` or `slice images` are available as **output type**. At the moment, `N5` or `slice images` are available as **output type**.
It is highly recommended to use N5 (and BDV or MoBIE for visualization) for large volumes or 2D slices. The slice export option can be chosen to produce downscaled images for talks or movie visualizations. It is highly recommended to use N5 (and BDV or MoBIE for visualization) for large volumes or 2D slices. The slice export option can be chosen to produce downscaled images for talks or movie visualizations.
![export](img/webui_export_computesettings.png "VolumeAlign WebUI export")
#### N5 #### N5
Pick the desired **output path** and the result will appear in a subdirectory `aligned` with the date and a reference to the stack. Pick the desired **output path** and the result will appear in a subdirectory `aligned` with the date and a reference to the stack.
...@@ -177,7 +192,7 @@ Make sure that the contrast in the slice viewer fits your expectations of the fi ...@@ -177,7 +192,7 @@ Make sure that the contrast in the slice viewer fits your expectations of the fi
With the known number of tiles and the estimated run time for exporting a slice, the necessary compute resources for the stack are predicted and can be checked in **Compute settings**. With the known number of tiles and the estimated run time for exporting a slice, the necessary compute resources for the stack are predicted and can be checked in **Compute settings**.
![export](img/webui_export_computesettings.png "VolumeAlign WebUI export")
Launching the computation will request the selected resources on the cluster and then launch a Spark instance on the allocated compute nodes that distributes and manages the parallel computation of the point matches. Launching the computation will request the selected resources on the cluster and then launch a Spark instance on the allocated compute nodes that distributes and manages the parallel computation of the point matches.
You will receive an email once the computation is done. It will tell you that the computation was `CANCELLED` but this only means that the resource allocation has been ended after successful computation of the matches. If you get a message referring to a `TIMEOUT`, you have to re-run the computation with more generous resource settings. You will receive an email once the computation is done. It will tell you that the computation was `CANCELLED` but this only means that the resource allocation has been ended after successful computation of the matches. If you get a message referring to a `TIMEOUT`, you have to re-run the computation with more generous resource settings.
...@@ -187,10 +202,11 @@ You will receive an email once the computation is done. It will tell you that th ...@@ -187,10 +202,11 @@ You will receive an email once the computation is done. It will tell you that th
This will export your slice(s) into the desired **output file type** which can be either `jpg`,`png`or `tif`. This will export your slice(s) into the desired **output file type** which can be either `jpg`,`png`or `tif`.
You can use **Choose output scale** to select the downscaling factor you like your export to have. The dimensions of the output images are shown in the status table in **Compute Settings**. There you can also choose the number of CPUs to use for parallel export. You can use **Choose output scale** to select the downscaling factor you like your export to have. The dimensions of the output images are shown in the status table in **Compute Settings**. There you can also choose the number of CPUs to use for parallel export.
## Finalize output format ## Finalize output format
Here you can finalize your data export (if `N5` was chosen for export) and generate the necessary metadata file to view the volume in BigDataViewer or MoBIE. Here you can finalize your data export (if `N5` was chosen for export) and generate the necessary metadata file to view the volume in BigDataViewer or MoBIE.
- **BDV-XML:** This will create the `XML`file necessary to view the result in *BigDataViewer* - **BDV-XML:** This will create the `XML`file necessary to view the result in *BigDataViewer*
- **MoBIE:** you can choose either an existing MoBIE project as output directory into which the aligned images will be added, or a new directory where the script will generate an entirely fresh MoBIE project. - **MoBIE:** you can choose either an existing MoBIE project as output directory into which the aligned images will be added, or a new directory where the script will generate an entirely fresh MoBIE project.
- **Choose exported volume:** This provides a list of already exported volumes. The files have to still reside in the original location on the storage system. - **Choose exported volume:** This provides a list of already exported volumes (in N5 format). The files have to still reside in the original location on the storage system.
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