Binary segmentation, zooming

[dolejm1]

Hi,
is it possible to visualize binary segmentation texture in any OMERO viewer?
I found an workaround by creating the new image with n+1 channels where the new channel is my binary segmentation, but this workflow copy the original data. So is it possible to combine two (or more) images without making the copy of them?

The second question (suggestion) is about zooming. Is it possible to zoom more than 100%? It is useful while annotating tiny objects (just one or two pixels wide).
Thank you,
Martin.

[wmoore]

Hi Martin,

I'm afraid it's not currently possible to view an overlay of one image on another in the OMERO viewers. The closest that I can think of is a 'webtest' prototype for overlaying channels from different images, E.g. in the screenshot below I'm showing the DAPI channel from 1 image overlaid on Red and Green channels from another image. This is available at <your-server>/webtest/channel_overlay_viewer/<imageId>. However, we have not yet tried to put this idea in our other clients.

The ability to zoom in beyond 100% certainly should be available in our clients. E.g. see the Insight screenshot below, zoomed in to 300%, using the vertical slider on the left. Do you not see this? Is 300% zoom enough, or do you need higher than this?

Regards,

Will.

[Trefex]

This is a really good question.

As a lot of image analysis is now done automatically, it would be great if one could visualize a binary mask over an existing image with an alpha blend.

Perhaps this could become a feature request?

[manics]

I've added a feature request to our ticketing system: https://trac.openmicroscopy.org.uk/ome/ticket/12660
Let me know if you want to be added to the CC list so you're notified of progress.

Simon

[Trefex]

Yes please

[i.munro]

Hi Martin

I'm not sure how applicable this is for you but it is possible to annotate an image with binary mask ROIs.
Further you can set the colour for each mask.

I attach a screenshot of the result viewed using OMERO.insight.
Here there are several segmentations all coloured semi-transparent red. Mostly the segmentations overlap
giving an opaque red but in the top-right of the image there are two regions of fainter red where there is no overlap.

I can point you at the Matlab code used to do this if it helps?

Regards

Ian