Skip to main content
Large-scale data management and sharing with Zarr and webKnossos
Norman Rzepka
Presenting author:
Norman Rzepka
Achieving interoperability and accessibility for very large image datasets is challenging, because the data cannot easily be copied or converted into different formats.
Zarr is an image data format that has been developed for streaming access and modern storage systems. Data that is remotely stored, e.g. in cloud storage or on an HTTP server, can be used with any Zarr-compatible software tool. The streaming-based approach makes very large datasets more accessible, because they do not need to be downloaded in full. Because Zarr is an open format, which is gaining support from more and more software tools, it enables interoperability.
webKnossos is an open-source web-based tool for visualization, annotation, and secure data management of large-scale 3D image datasets. The software allows collaborators to manage their shared data in one place and to work with even the largest images from a web-based interface. We now release Zarr support for webKnossos. Not only can remote datasets be used within webKnossos, the software also exports its own data using the Zarr format. Therefore, data that is stored and managed by webKnossos can be used in any Zarr-compatible tool without copying the entire data. With the new Zarr features, webKnossos becomes a secure hub for storing data while strengthening interoperability with other analysis tools.