LOBSTER as community partner

I would like to announce the release of LOBSTER, my new BIA software:

LOBSTER (Little OBjects Segmentation and Tracking EnviRonment) is a modular MATLAB environment to design workflows characterizing biological objects in large microscopy images, and exporting object models in standard formats compatible with other open source projects. Especially, LOBSTER provides a tangible solution to the tracing of blood vessels and to the tracking of nuclei from 3D microscopy scans.


It would be great if LOBSTER could becomes a Community Partners, and I would then adopt image.sc as the primary forum for technical discussions regarding the software.


And please notice that in the previous community partner post they wanted the link to the forum before adding someone as a member.

I will add it straight away once you provide me the URL to the wepage on image.sc (should I assume forum.image.sc/tag/lobster?). Thanks!

I believe you can make it whatever you want it to be, based on the tag in your post. Though you added a lot of tags in your post :slight_smile:

I have no control over that, though, just making sure as many things are in order as possible when someone else notices this!

1 Like

I added a link to image.sc forum from LOBSTER documentation portal. Could you please add logo and side bar information as provided in my initial message?

1 Like

I don’t think I can, but I’m sure someone will take care of it soon!

Ok, thanks a lot for your help and for welcoming LOBSTER!

@Sebastien I added LOBSTER as community partner. Your tag sidebar topic is here.

Do you have a version of the logo that works on a dark background? Or are you OK with it being a white square for the dark theme?


Thanks Curtis! I am fine with the current logo for the dark theme.


can it deal with bright field images on Tracking dividing nuclei from a 2D time-lapse? Or just fluorescence?


LOBSTER main focus is labeled objects (typically fluorescence), but some techniques for image restoration are provided (phase contrast). If you can only use brightfield, state of the art nowadays is to use Deep learning (e.g. U-NET) to infere nuclei / cells, then use a tracker for labeled objects.