Combining segmentations after proof-reading


I use webKnossos to inspect my segmentations and fix errors in the over segmentation. For this proofreading I use the webKnossos merger mode tool. That works well and I can quickly combine several segments into a larger segments and fix splits.

My question is: How can I use this corrected version for further analysis? When I download the dataset, the segmentation layer is still my original over-segmentation and not the combined/proofread version I created.


Hi @MikeJacob,

I am happy to hear that the merger mode works well for you! In order to apply the merge operations on your data, you can download the annotation file and use the Python-based libraries to merge the segments. A few months ago, I wrote a sample script that does exactly that: · GitHub This script will return a sequence of tiff images.

There is also a version that produces a WKW dataset that you can re-import into webKnossos: · GitHub

We are working on a feature to apply the merging when using the data download. Stay tuned.

Thank you for the quick reply. I will try to get the scripts to run on my local machine and merge the data. I see this uses a wknml and wkw Python library for working with the webKnossos file format. Good to know.