Repository navigation
Reconstructed Image Overlay #378
Description
Part of #373 Live plotting:
Create the pipeline connection to automatically fetch reconstructed ptychography images and overlay them on the STXM scan canvas.
The reference frame is addressed in #379
Goals
Acceptance criteria : What journeys should the user be able to complete to consider your feature done
Be as specific as you can, check them off once they work. Add more if you find more requirements during development. Issue should only be considered done after all of them are checked
- Reconstructed images from the Nathanael's processing pipeline can be toggled on/off directly over the corresponding raw STXM scan layer
Tasks
base on the goals, deduce the tasks required. check them off once they're done
- Processing Pipeline Fetcher: Build an API client to listen for and pull completed or intermediate reconstructed ptychography outputs.
- Canvas Registration Hook: Dynamically append the fetched reconstructed image as a new registered layer in the multi-layer canvas stack.
Activity
[-]Integrate reconstructed image fetching to overlay processed ptychography outputs onto the canvas.[/-][+]Reconstructed Image Overlay[/+]DanPorter commentedon Sep 15, 2026
This is not just combining two images but potentially multiple STXM images and reconstructions at the same time, overlaying and tiling the images in a single plot. There will also be coordinate transformations to apply to each layer.
Relm-Arrowny commentedon Sep 15, 2026
Well sported, this need to be addressed in #379. As long as the data are serving the images in the corrected reference frame , the UI just has to stack and display the layers without needing to know about the transformation at all, is my current thinking.
NKatti2011 commentedon Sep 16, 2026
Below include notes from meeting with @EmsArnold and a meeting with David Burn.
A reconstruction covers the whole ROI, not one image per scan point. The overlay
therefore registers a single reconstructed image against the ROI's extent in the chosen
reference frame, rather than compositing a grid of per-point images. David expects users
will want to see the reconstructed images first and foremost.
Users are also expected to supply offsets for reconstructed images after processing,
in order to locate the ROI. This is being called the "alignment" stage, and it uses the
same transformation maths as #379, which makes the processing pipeline a third consumer
of that package, alongside controls and the UI.
Data
Status and change notifications
That is still in progress, so initially we connect to the Workflows websockets
directly and subscribe there.
their status, and submit a workflow when a scan ends:
ProcessedSpectroscopyData.tsxSpectroscopyView.tsxwebsocket when it lands, without touching the components.
teams are using as their client, and consistency is good.
To scope
turns out to be, it may be better done in the Davidia plot server instead.