Objects And Overlays¶
odon can display more than image channels. You can layer objects, labels, points, and masks on top of the image canvas.
Common Overlay Types¶
- segmentation outlines from NGFF labels
- polygon or point objects from GeoParquet, Parquet, GeoJSON, or CSV
- exclusion or artefact masks from GeoJSON
- point overlays such as transcript-like data
Object Data¶
For object-centric review, the most useful inputs are usually:
- GeoParquet for rich polygon or point data with scalar properties
- GeoJSON for simple interchange
- CSV for point-only data with
xandycolumns
These layers can participate in selection, inspection, and some analysis workflows.
NGFF Labels¶
If the dataset includes label groups under labels/, the viewer can render outlines for those label layers.
Useful notes:
- label groups are discoverable rather than assumed to be called
cells - the viewer renders outlines rather than filled masks
- label appearance can be adjusted from the layer properties
Masks¶
The viewer supports exclusion or artefact masks saved as GeoJSON.
Current workflow:
- load masks from the
ROI Selectortab - draw or edit masks using the
Draw masktool - export masks to GeoJSON when needed
Points¶
Point overlays are available through both direct point loading and specialized dataset adapters.
Examples include:
- generic point tables
- thresholded cell-centroid style views
- Xenium transcript-style overlays
Recommended Overlay Workflow¶
- Open the image dataset first.
- Add segmentation or object layers.
- Select the active overlay layer in the left panel.
- Use the right panel to adjust appearance or analysis settings.
- Keep image channels beneath non-image overlays for clarity.